From d873384dddfee0e5a48a553f8c6185e644022d92 Mon Sep 17 00:00:00 2001 From: yi-ge Date: Tue, 13 Oct 2026 11:37:59 +0000 Subject: [PATCH] =?UTF-8?q?=E5=8F=91=E5=B8=83=E9=85=8D=E7=BD=AE=E4=B8=AD?= =?UTF-8?q?=E5=BF=83=E8=BF=90=E8=A1=8C=E6=95=B0=E6=8D=AE=20131?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Source-Commit: d77e4a27acf825961164ad1762d575264ee554a8 Payload-SHA256: e7abe73a4c9477047dc4a087ed981d687bd3c70d8cb3ae76b626784f8dcae141 --- .desirecore-runtime.json | 1 + api-providers/image-understanding/_index.json | 4 + .../configured-vision-anthropic.json | 32 + .../configured-vision-openai.json | 30 + api-providers/web-search/_index.json | 4 + api-providers/web-search/brave.json | 35 + .../web-search/doubao-search-global.json | 36 + api-providers/web-search/serper.json | 19 + api-providers/web-search/tavily.json | 23 + catalog/v2/retirements.json | 24 + compute/coding-plans/_index.json | 16 + compute/coding-plans/baidu-coding.json | 178 + compute/coding-plans/dashscope-coding.json | 210 + .../coding-plans/dashscope-token-plan.json | 549 + compute/coding-plans/infini-coding.json | 43 + compute/coding-plans/kwai-coding.json | 38 + compute/coding-plans/minimax-coding.json | 226 + compute/coding-plans/moonshot-coding.json | 33 + compute/coding-plans/moorethread-coding.json | 38 + compute/coding-plans/tencent-token.json | 180 + compute/coding-plans/volcengine-coding.json | 259 + compute/coding-plans/zhipu-coding.json | 197 + compute/model-specs/_index.json | 63 + compute/model-specs/anthropic.json | 419 + compute/model-specs/baichuan.json | 111 + compute/model-specs/baidu.json | 146 + compute/model-specs/cohere.json | 297 + compute/model-specs/deepseek.json | 354 + compute/model-specs/google.json | 535 + compute/model-specs/happyhorse.json | 184 + compute/model-specs/kling.json | 133 + compute/model-specs/minimax.json | 762 + compute/model-specs/mistral.json | 107 + compute/model-specs/moonshot.json | 255 + compute/model-specs/openai.json | 1324 + compute/model-specs/perplexity.json | 85 + compute/model-specs/qwen.json | 1188 + compute/model-specs/stability.json | 50 + compute/model-specs/stealth.json | 46 + compute/model-specs/tencent.json | 265 + compute/model-specs/volcengine.json | 808 + compute/model-specs/wan.json | 106 + compute/model-specs/xai.json | 155 + compute/model-specs/xiaomi.json | 556 + compute/model-specs/xunfei.json | 61 + compute/model-specs/zhipu.json | 462 + compute/pricing.json | 4 + compute/providers/_index.json | 34 + compute/providers/anthropic-claude.json | 239 + compute/providers/anthropic.json | 311 + compute/providers/baichuan.json | 107 + compute/providers/baidu.json | 158 + compute/providers/cohere.json | 151 + compute/providers/dashscope.json | 943 + compute/providers/deepseek.json | 219 + compute/providers/github-copilot.json | 28 + compute/providers/google.json | 323 + compute/providers/kling.json | 104 + compute/providers/local-whisper.json | 38 + compute/providers/minimax.json | 728 + compute/providers/mistral.json | 125 + compute/providers/moonshot.json | 419 + compute/providers/ollama.json | 34 + compute/providers/openai-codex.json | 347 + compute/providers/openai.json | 1068 + compute/providers/openrouter.json | 318 + compute/providers/perplexity.json | 95 + compute/providers/siliconflow.json | 80 + compute/providers/stability.json | 43 + compute/providers/tencent.json | 180 + compute/providers/volcengine.json | 712 + compute/providers/xai.json | 143 + compute/providers/xiaomi.json | 248 + compute/providers/xunfei-x2.json | 41 + compute/providers/xunfei.json | 42 + compute/providers/zhipu-embedding.json | 39 + compute/providers/zhipu.json | 556 + compute/service-map.json | 82 + compute/v2/catalog.json | 30343 ++++++++++++++++ manifest.json | 6 + runtimes/recommended.json | 44 + runtimes/versions-fallback.json | 1437 + schemas/model-spec.schema.json | 311 + schemas/model-specs-index.schema.json | 88 + 84 files changed, 50835 insertions(+) create mode 100644 .desirecore-runtime.json create mode 100644 api-providers/image-understanding/_index.json create mode 100644 api-providers/image-understanding/configured-vision-anthropic.json create mode 100644 api-providers/image-understanding/configured-vision-openai.json create mode 100644 api-providers/web-search/_index.json create mode 100644 api-providers/web-search/brave.json create mode 100644 api-providers/web-search/doubao-search-global.json create mode 100644 api-providers/web-search/serper.json create mode 100644 api-providers/web-search/tavily.json create mode 100644 catalog/v2/retirements.json create mode 100644 compute/coding-plans/_index.json create mode 100644 compute/coding-plans/baidu-coding.json create mode 100644 compute/coding-plans/dashscope-coding.json create mode 100644 compute/coding-plans/dashscope-token-plan.json create mode 100644 compute/coding-plans/infini-coding.json create mode 100644 compute/coding-plans/kwai-coding.json create mode 100644 compute/coding-plans/minimax-coding.json create mode 100644 compute/coding-plans/moonshot-coding.json create mode 100644 compute/coding-plans/moorethread-coding.json create mode 100644 compute/coding-plans/tencent-token.json create mode 100644 compute/coding-plans/volcengine-coding.json create mode 100644 compute/coding-plans/zhipu-coding.json create mode 100644 compute/model-specs/_index.json create mode 100644 compute/model-specs/anthropic.json create mode 100644 compute/model-specs/baichuan.json create mode 100644 compute/model-specs/baidu.json create mode 100644 compute/model-specs/cohere.json create mode 100644 compute/model-specs/deepseek.json create mode 100644 compute/model-specs/google.json create mode 100644 compute/model-specs/happyhorse.json create mode 100644 compute/model-specs/kling.json create mode 100644 compute/model-specs/minimax.json create mode 100644 compute/model-specs/mistral.json create mode 100644 compute/model-specs/moonshot.json create mode 100644 compute/model-specs/openai.json create mode 100644 compute/model-specs/perplexity.json create mode 100644 compute/model-specs/qwen.json create mode 100644 compute/model-specs/stability.json create mode 100644 compute/model-specs/stealth.json create mode 100644 compute/model-specs/tencent.json create mode 100644 compute/model-specs/volcengine.json create mode 100644 compute/model-specs/wan.json create mode 100644 compute/model-specs/xai.json create mode 100644 compute/model-specs/xiaomi.json create mode 100644 compute/model-specs/xunfei.json create mode 100644 compute/model-specs/zhipu.json create mode 100644 compute/pricing.json create mode 100644 compute/providers/_index.json create mode 100644 compute/providers/anthropic-claude.json create mode 100644 compute/providers/anthropic.json create mode 100644 compute/providers/baichuan.json create mode 100644 compute/providers/baidu.json create mode 100644 compute/providers/cohere.json create mode 100644 compute/providers/dashscope.json create mode 100644 compute/providers/deepseek.json create mode 100644 compute/providers/github-copilot.json create mode 100644 compute/providers/google.json create mode 100644 compute/providers/kling.json create mode 100644 compute/providers/local-whisper.json create mode 100644 compute/providers/minimax.json create mode 100644 compute/providers/mistral.json create mode 100644 compute/providers/moonshot.json create mode 100644 compute/providers/ollama.json create mode 100644 compute/providers/openai-codex.json create mode 100644 compute/providers/openai.json create mode 100644 compute/providers/openrouter.json create mode 100644 compute/providers/perplexity.json create mode 100644 compute/providers/siliconflow.json create mode 100644 compute/providers/stability.json create mode 100644 compute/providers/tencent.json create mode 100644 compute/providers/volcengine.json create mode 100644 compute/providers/xai.json create mode 100644 compute/providers/xiaomi.json create mode 100644 compute/providers/xunfei-x2.json create mode 100644 compute/providers/xunfei.json create mode 100644 compute/providers/zhipu-embedding.json create mode 100644 compute/providers/zhipu.json create mode 100644 compute/service-map.json create mode 100644 compute/v2/catalog.json create mode 100644 manifest.json create mode 100644 runtimes/recommended.json create mode 100644 runtimes/versions-fallback.json create mode 100644 schemas/model-spec.schema.json create mode 100644 schemas/model-specs-index.schema.json diff --git a/.desirecore-runtime.json b/.desirecore-runtime.json new file mode 100644 index 0000000..967a50c --- /dev/null +++ b/.desirecore-runtime.json @@ -0,0 +1 @@ +{"formatVersion":1,"profile":"config-center-runtime-v1","dataVersion":131,"payloadDigest":"e7abe73a4c9477047dc4a087ed981d687bd3c70d8cb3ae76b626784f8dcae141","contracts":{"modelSpecRoutingSha256":"a413beb0819af93e79eaf4da53bfb4b723717c1aa143e7c7a9c6f7644d278e68","modelSpecsIndexSha256":"0c769bb2f357b717b3300027d8009cba2dd61b031814f7a1519e4dba66a3560b"},"files":[{"path":"api-providers/image-understanding/_index.json","byteLength":146,"sha256":"ef6095d2260de331e0d2e847be008dd5e2a45516dbf5c9432a73b12e5a003641"},{"path":"api-providers/image-understanding/configured-vision-anthropic.json","byteLength":935,"sha256":"c3bc4ef943974b58c176d45381177c93bcad4b69637e0ce5ccd1426df675de63"},{"path":"api-providers/image-understanding/configured-vision-openai.json","byteLength":793,"sha256":"f0359ffd6900d5c296927e636e7a74539258a888a0b2643b3cfe45aef0aa6133"},{"path":"api-providers/web-search/_index.json","byteLength":131,"sha256":"8cf6c69db16083c0303ba783a8a762f76949024423573fe26a6cb4610bb14756"},{"path":"api-providers/web-search/brave.json","byteLength":736,"sha256":"d5385151585a0acf4721fb444f80735736d2521be02ae184d60258d81e89067d"},{"path":"api-providers/web-search/doubao-search-global.json","byteLength":817,"sha256":"4db70b46637fc07e4cb55855626c1f6ae2848ffd453cb8cf3ca09bb1d8a652d5"},{"path":"api-providers/web-search/serper.json","byteLength":519,"sha256":"1b9acd975ab88a15758775d205690b1723fbca381eba133fd2668689b652f830"},{"path":"api-providers/web-search/tavily.json","byteLength":547,"sha256":"709709907a663f9820c4ffe00b006fce374aef7b9bc2923ef1c17a09ad0f0111"},{"path":"catalog/v2/retirements.json","byteLength":939,"sha256":"3387a5c84de8a8b0f46c384b97c0b956555b6d206d144c15b30f35257b5103bb"},{"path":"compute/coding-plans/_index.json","byteLength":339,"sha256":"307ad2862f3457b17e5a95b8e1f2d7553e11f718cbf8f12035e2f156065bf26c"},{"path":"compute/coding-plans/baidu-coding.json","byteLength":4107,"sha256":"af3bf397b56378eba9457ee0ca2f0cdebe69db8b23d8009c9f178bf7ad37ecd8"},{"path":"compute/coding-plans/dashscope-coding.json","byteLength":5039,"sha256":"aedcbd74853067c7d05670a169a2355637bad942f71a6d25dbe302b6e2a40b4b"},{"path":"compute/coding-plans/dashscope-token-plan.json","byteLength":13433,"sha256":"0bc3fdd62f033eaa26bf3157975214678115641b25d572b99885ff265c958054"},{"path":"compute/coding-plans/infini-coding.json","byteLength":996,"sha256":"f6b9b6d9be03da3e5997c154dacd06cc033a226b11276ee935050c1945079b17"},{"path":"compute/coding-plans/kwai-coding.json","byteLength":874,"sha256":"64c936620fabc7880b80b38572e1a26bd5424bb53eebb4fff3ba384e6d4c8162"},{"path":"compute/coding-plans/minimax-coding.json","byteLength":5963,"sha256":"820f474379f18b363c0ce88d52fdde409bbab08a6b848de279fdb3e551c8b82b"},{"path":"compute/coding-plans/moonshot-coding.json","byteLength":969,"sha256":"7695e0c93e9c19c83c94f84873d82ea4c88aac40733b4cd6751e1c2c20266318"},{"path":"compute/coding-plans/moorethread-coding.json","byteLength":849,"sha256":"baca56618645a0e2ebaab0fd146eef3c6e48d0118aa14173f48be87225e2bee3"},{"path":"compute/coding-plans/tencent-token.json","byteLength":3999,"sha256":"ea11276c169d76d94719b9c16fca1a8d381e273fe1503987b47bdb5ff651c2fc"},{"path":"compute/coding-plans/volcengine-coding.json","byteLength":6211,"sha256":"bee80d67e37a3c5cd2e07e52534a80979bca514e3bea641bfc8d640eba17eff8"},{"path":"compute/coding-plans/zhipu-coding.json","byteLength":4879,"sha256":"9be189a2fe6806328f43b2d601bf70392e41004ceb8a865fcee01aa2da7f4031"},{"path":"compute/model-specs/_index.json","byteLength":1898,"sha256":"89defaa701033e3fb0911c5984032c027ea6b39956e7cd70440d11cbc40485f4"},{"path":"compute/model-specs/anthropic.json","byteLength":10161,"sha256":"cc0d20633dd30ae64f74ab890b0a0c7a18ff3bd9b843fad9cea56390d48c536d"},{"path":"compute/model-specs/baichuan.json","byteLength":2643,"sha256":"2d92c8fdf33782a713f0f86171f5e568dc680c893e8493c089000ddd842d7aec"},{"path":"compute/model-specs/baidu.json","byteLength":3612,"sha256":"11d000f4454cfd7323e93a90e7a108b2ada2d994d510e3e552698f0cd9ba4429"},{"path":"compute/model-specs/cohere.json","byteLength":6799,"sha256":"f0f4a10997030a053f954d6ece055529cb964a05054f7f0195343c711971bc5a"},{"path":"compute/model-specs/deepseek.json","byteLength":8390,"sha256":"f9f21146e6eb8e36c85d18b328bb757f1b8c8dcccc73ce11ea550ea6109204ca"},{"path":"compute/model-specs/google.json","byteLength":13030,"sha256":"691b98d590735ae4e5069ddfe12a15cc71e8770ec6a451ed15af7729c1d56f05"},{"path":"compute/model-specs/happyhorse.json","byteLength":4610,"sha256":"2b46b65f17844baf07104e83dcc4eedebba3eee07dc196f9c236ae30f0b72973"},{"path":"compute/model-specs/kling.json","byteLength":3137,"sha256":"8967d26fa0151914eb07446e29df3e03a298d598f9972949ece9cc93cbb35872"},{"path":"compute/model-specs/minimax.json","byteLength":18472,"sha256":"a6eb1031e1a5a151bbfa9019665271e53c1f9b1f641b5a5ff77c2217df3d582a"},{"path":"compute/model-specs/mistral.json","byteLength":2375,"sha256":"b0e51cfd8cb868d323d7bab0689c8d4fb03d411a9402c8da37a551eea552b554"},{"path":"compute/model-specs/moonshot.json","byteLength":5917,"sha256":"6bcf0cd0e3fd7ed3d2813b81a08f31ddd4f5d95958eb7c059d257bf77b0f849f"},{"path":"compute/model-specs/openai.json","byteLength":30216,"sha256":"c902249ae9aa240f16469890ecde387c81b3d8f00f2fd9080cb3a7219e64b78f"},{"path":"compute/model-specs/perplexity.json","byteLength":2014,"sha256":"d7ca6e43fb6d4b729e69619689c60ed03d2ad24ceba3378f1923a38762e91ffc"},{"path":"compute/model-specs/qwen.json","byteLength":27471,"sha256":"e0b032fd39615295d31ad2703bc7043365dcc5dd0efafbf7d54d37306976040e"},{"path":"compute/model-specs/stability.json","byteLength":1344,"sha256":"bdad05b103ec4ee61f10d791503719efc30ee12b8b909f22bd9fdba36467bd6a"},{"path":"compute/model-specs/stealth.json","byteLength":1446,"sha256":"ecd25432d1637c5b837d3adaadd25caa49ed06553426eac94b3862132e89fadd"},{"path":"compute/model-specs/tencent.json","byteLength":6413,"sha256":"f381358f5ff4e99affa5e7bd37d8a2938ac6997a6e72669ff72562ec0f0fe03d"},{"path":"compute/model-specs/volcengine.json","byteLength":19911,"sha256":"1f3055a32383aaf8cf844a8acdb28902a31e6442fe4f51c631b61aceb847d30e"},{"path":"compute/model-specs/wan.json","byteLength":2785,"sha256":"356331e8c91577fad0650210e5cabee956e22e6fe1e802621ae1ff21d1287d8a"},{"path":"compute/model-specs/xai.json","byteLength":3604,"sha256":"a3316bbe4bb3ebde1fbd4ede0c73db57d6dd76752f642607c3b4145bb6730ef0"},{"path":"compute/model-specs/xiaomi.json","byteLength":13609,"sha256":"50628803583cc49549c2baa480737a97be8f276969e3fa17a63e940a51dcd68f"},{"path":"compute/model-specs/xunfei.json","byteLength":1538,"sha256":"6e7ad3def089fb5589ab012766de161df1a84360266ed73bbc25acc711986fc1"},{"path":"compute/model-specs/zhipu.json","byteLength":11032,"sha256":"0726e0f31d2652a0de393611252057ea05dd929243c5172cf441d8065dc8a892"},{"path":"compute/pricing.json","byteLength":44,"sha256":"e94a6d322755a9ab3841219bb043cdb0fe8e2b3b9b41d9b76ad582555a8d5833"},{"path":"compute/providers/_index.json","byteLength":546,"sha256":"b15dd80c98088f0b2e233124422a4b285a9ad7f301a23da000017b60df51d8fb"},{"path":"compute/providers/anthropic-claude.json","byteLength":6171,"sha256":"f35cc85af4f0032bed8927ab3b4f25a86da2a871c92dd84698a42a52ca71bd32"},{"path":"compute/providers/anthropic.json","byteLength":8513,"sha256":"88c81af0550c6a202829d7e70ec675e56f9e11f58edaa62736dbfdd2300dafbf"},{"path":"compute/providers/baichuan.json","byteLength":2561,"sha256":"104f7d97ba53c349de63f05c0caf621638e560d0f4f8360f0343dc39a7e11c3c"},{"path":"compute/providers/baidu.json","byteLength":3854,"sha256":"108196a0e6d13935f9b3a6607781320e084fbd5e55181e4507914ad53c6e1f69"},{"path":"compute/providers/cohere.json","byteLength":4220,"sha256":"4af001dcc2d283d93de6aa5e7e6b9695f63cc6e82ab448005b3f9ce4e2d401fc"},{"path":"compute/providers/dashscope.json","byteLength":24564,"sha256":"7710b8305f4fe72644116477ba6717c8ac7ea8269f72fa470b10a9b905d86d02"},{"path":"compute/providers/deepseek.json","byteLength":5765,"sha256":"e1c99b895b0aed243e4defd73e8e760ba04bfadbbdbf9ac11064922f81a41f4f"},{"path":"compute/providers/github-copilot.json","byteLength":708,"sha256":"132181d570923a0dd99a9097b2a8ef42312757d1e02eda82c30c631cc434ce13"},{"path":"compute/providers/google.json","byteLength":8509,"sha256":"711a8ea1fe9036c535d201ed19636ac01309473e2802cae09a197c97514d97b9"},{"path":"compute/providers/kling.json","byteLength":2485,"sha256":"3cce75260edeea3a42530532950c8ea7ddff1ce4ba4946f169b3a15dd9bfe3e3"},{"path":"compute/providers/local-whisper.json","byteLength":817,"sha256":"f85496b997a8f0bb3955eb2c76b678fac9b691b03481c6cf74ce294c78e2b358"},{"path":"compute/providers/minimax.json","byteLength":18180,"sha256":"11314a2a75298cd053bd913b11a6d7e71b9cfb2b7ff8ba2ec7bf6db7547dea4c"},{"path":"compute/providers/mistral.json","byteLength":3009,"sha256":"0f5eb09641f108b609918e2478f6aa9a568816fe189a5abd3ad85e5e39770a2e"},{"path":"compute/providers/moonshot.json","byteLength":10280,"sha256":"848c7246b942eb1b5b84af475b35c69e695704f5b52eb9c29b5f110cf33c6a03"},{"path":"compute/providers/ollama.json","byteLength":752,"sha256":"5fc29731f6857c638e0612fcbce805d42e28f5b5b393ce4e20247026746be784"},{"path":"compute/providers/openai-codex.json","byteLength":8510,"sha256":"ca45b6ac74be1dfda5107cee038329747144c119d47833c8257fc3234caa857b"},{"path":"compute/providers/openai.json","byteLength":27287,"sha256":"02b9bfaa872b85c36270e087edecbb27e6d229ae2410721b347d2ff8b48a429c"},{"path":"compute/providers/openrouter.json","byteLength":8583,"sha256":"6f37b09b338fc806bf2ef28867b516f1d15333417f7c376147314b237af64a05"},{"path":"compute/providers/perplexity.json","byteLength":2435,"sha256":"91b6a29b6b017870a96f422b1909565fa2888f840b902ddacbcf8b862cbb3a2c"},{"path":"compute/providers/siliconflow.json","byteLength":2063,"sha256":"365b685e9c5786ca405bd65d538be83df3a2775ffcface1dcc3651fcbce6a22a"},{"path":"compute/providers/stability.json","byteLength":989,"sha256":"32e096a11412cfc752700ecb77457117e4474e9dc3bb8b668c1dd13d7b5c366c"},{"path":"compute/providers/tencent.json","byteLength":4468,"sha256":"89e8c0d749ced86324184e2fb2e22a9e8e1c8ab978a056edd7bc2d6e5f0b0d6b"},{"path":"compute/providers/volcengine.json","byteLength":18931,"sha256":"b9aa1bfdee6c949db1cb0f2a9fd9f2a24276dc7cd020078d8805d2f397215e26"},{"path":"compute/providers/xai.json","byteLength":3482,"sha256":"02d2a42ce6632bb48fc77894a67f8ac2ff975a085f0b66199c4eb7380bd1f27b"},{"path":"compute/providers/xiaomi.json","byteLength":6494,"sha256":"da62f3136aa139517251ea14dcbeb2a128e3115be90feb4fa8436ab53ec7300a"},{"path":"compute/providers/xunfei-x2.json","byteLength":1048,"sha256":"0bdfdf053d3f2a7cd91dcbe389998ea55dfa00e11ff50b6f6300463bc06067ca"},{"path":"compute/providers/xunfei.json","byteLength":1013,"sha256":"0b6a665bafe2e4f72304b699f06f83123ba4d32d9d9f90364680fde470ac8d9e"},{"path":"compute/providers/zhipu-embedding.json","byteLength":1069,"sha256":"3b14bf0a5be8c6cd839d7657890624458b703c882efc9bca34443d0afd63a1f5"},{"path":"compute/providers/zhipu.json","byteLength":14417,"sha256":"5fd1d0344eeb64464d564ddc720a27c7de3d4acf0f80394e59fc26518f9e6a64"},{"path":"compute/service-map.json","byteLength":1937,"sha256":"4c7734ea228f727157f4251c5abc249777dda8683e6326e99cd1abd41163692a"},{"path":"compute/v2/catalog.json","byteLength":944125,"sha256":"744dd72f5882bb5b39374b5a81ea2e11b487736b8818b2eed081da9c717bec77"},{"path":"manifest.json","byteLength":132,"sha256":"5c530bd58ce63cf55a3d57d9463dc61e7bf06815ab4dd5e4bfc209ed037c4c12"},{"path":"runtimes/recommended.json","byteLength":2380,"sha256":"5d06494dbcd40eabc1425099f52414fcc16c0cf3b8df1bf1e350b52ef5e8d1dc"},{"path":"runtimes/versions-fallback.json","byteLength":31229,"sha256":"6ce53f2e2f37c9c413eb5627f9cd1bdbc9c2268d19421f26f72af5c2acc4db8d"},{"path":"schemas/model-spec.schema.json","byteLength":18666,"sha256":"1169df90078451d0e7d68cd7a2a986923441f39909a0c6e0f2e8b530887c78ce"},{"path":"schemas/model-specs-index.schema.json","byteLength":3217,"sha256":"728b350de9def0112e5b7db87397eaac6009d9f0f5c51dbfeeca5f881a0378eb"}]} diff --git a/api-providers/image-understanding/_index.json b/api-providers/image-understanding/_index.json new file mode 100644 index 0000000..f576e3d --- /dev/null +++ b/api-providers/image-understanding/_index.json @@ -0,0 +1,4 @@ +{ + "description": "image_understanding 工具后端加载顺序索引", + "order": ["configured-vision-openai", "configured-vision-anthropic"] +} diff --git a/api-providers/image-understanding/configured-vision-anthropic.json b/api-providers/image-understanding/configured-vision-anthropic.json new file mode 100644 index 0000000..0d47d02 --- /dev/null +++ b/api-providers/image-understanding/configured-vision-anthropic.json @@ -0,0 +1,32 @@ +{ + "id": "configured-vision-anthropic", + "name": "视觉模型（Anthropic 兼容，复用默认映射）", + "capability": "image_understanding", + "enabled": true, + "builtin": true, + "priority": 20, + "endpoint": "${baseUrl}/messages", + "method": "POST", + "computeServiceType": "vision", + "requiresApiFormat": "anthropic-messages", + "auth": { "type": "header", "headerName": "x-api-key" }, + "request": { + "imageMode": "base64", + "headers": { "anthropic-version": "2023-06-01" }, + "bodyTemplate": { + "model": "${model}", + "max_tokens": 1024, + "messages": [ + { + "role": "user", + "content": [ + { "type": "image", "source": { "type": "base64", "media_type": "${imageMime}", "data": "${imageBase64}" } }, + { "type": "text", "text": "${prompt}" } + ] + } + ] + } + }, + "response": { "textPath": "content.0.text" }, + "timeoutMs": 30000 +} diff --git a/api-providers/image-understanding/configured-vision-openai.json b/api-providers/image-understanding/configured-vision-openai.json new file mode 100644 index 0000000..7e453a4 --- /dev/null +++ b/api-providers/image-understanding/configured-vision-openai.json @@ -0,0 +1,30 @@ +{ + "id": "configured-vision-openai", + "name": "视觉模型（OpenAI 兼容，复用默认映射）", + "capability": "image_understanding", + "enabled": true, + "builtin": true, + "priority": 10, + "endpoint": "${baseUrl}/chat/completions", + "method": "POST", + "computeServiceType": "vision", + "requiresApiFormat": "openai-completions", + "auth": { "type": "bearer" }, + "request": { + "imageMode": "url", + "bodyTemplate": { + "model": "${model}", + "messages": [ + { + "role": "user", + "content": [ + { "type": "text", "text": "${prompt}" }, + { "type": "image_url", "image_url": { "url": "${imageUrl}" } } + ] + } + ] + } + }, + "response": { "textPath": "choices.0.message.content" }, + "timeoutMs": 30000 +} diff --git a/api-providers/web-search/_index.json b/api-providers/web-search/_index.json new file mode 100644 index 0000000..02cca3d --- /dev/null +++ b/api-providers/web-search/_index.json @@ -0,0 +1,4 @@ +{ + "description": "web_search 工具后端加载顺序索引", + "order": ["tavily", "brave", "serper", "doubao-search-global"] +} diff --git a/api-providers/web-search/brave.json b/api-providers/web-search/brave.json new file mode 100644 index 0000000..a3389e1 --- /dev/null +++ b/api-providers/web-search/brave.json @@ -0,0 +1,35 @@ +{ + "id": "brave", + "name": "Brave Search API", + "capability": "web_search", + "enabled": false, + "builtin": true, + "priority": 20, + "endpoint": "https://api.search.brave.com/res/v1/web/search", + "method": "GET", + "auth": { + "type": "header", + "headerName": "X-Subscription-Token", + "apiKeyRef": "brave" + }, + "request": { + "headers": { + "Accept": "application/json", + "Accept-Encoding": "gzip" + }, + "queryParams": { + "q": "${query}", + "count": "${maxResults}" + } + }, + "response": { + "resultsPath": "web.results", + "item": { + "titlePath": "title", + "urlPath": "url", + "snippetPath": "description", + "pageAgePath": "page_age" + } + }, + "timeoutMs": 10000 +} diff --git a/api-providers/web-search/doubao-search-global.json b/api-providers/web-search/doubao-search-global.json new file mode 100644 index 0000000..2f79ea0 --- /dev/null +++ b/api-providers/web-search/doubao-search-global.json @@ -0,0 +1,36 @@ +{ + "id": "doubao-search-global", + "name": "豆包搜索 Global", + "capability": "web_search", + "enabled": false, + "builtin": true, + "priority": 5, + "endpoint": "https://open.feedcoopapi.com/search_api/global_search", + "method": "POST", + "auth": { + "type": "bearer", + "apiKeyRef": "doubao-search-global" + }, + "request": { + "bodyTemplate": { + "Query": "${query}", + "SearchType": "web", + "DocCount": "${maxResults}", + "MaxSnippetLength": 500 + } + }, + "response": { + "resultsPath": "Result.Documents", + "item": { + "titlePath": "Title", + "urlPath": "Url", + "snippetPath": "Snippet.0.Text", + "pageAgePath": "DocumentInfo.PublishTime" + }, + "successCondition": { + "path": "Result.ErrorCode", + "equals": 0 + } + }, + "timeoutMs": 10000 +} diff --git a/api-providers/web-search/serper.json b/api-providers/web-search/serper.json new file mode 100644 index 0000000..c2c781c --- /dev/null +++ b/api-providers/web-search/serper.json @@ -0,0 +1,19 @@ +{ + "id": "serper", + "name": "Serper", + "capability": "web_search", + "enabled": false, + "builtin": true, + "priority": 30, + "endpoint": "https://google.serper.dev/search", + "method": "POST", + "auth": { "type": "header", "headerName": "X-API-KEY", "apiKeyRef": "serper" }, + "request": { + "bodyTemplate": { "q": "${query}", "num": "${maxResults}" } + }, + "response": { + "resultsPath": "organic", + "item": { "titlePath": "title", "urlPath": "link", "snippetPath": "snippet" } + }, + "timeoutMs": 10000 +} diff --git a/api-providers/web-search/tavily.json b/api-providers/web-search/tavily.json new file mode 100644 index 0000000..9f37618 --- /dev/null +++ b/api-providers/web-search/tavily.json @@ -0,0 +1,23 @@ +{ + "id": "tavily", + "name": "Tavily", + "capability": "web_search", + "enabled": false, + "builtin": true, + "priority": 10, + "endpoint": "https://api.tavily.com/search", + "method": "POST", + "auth": { "type": "bearer", "apiKeyRef": "tavily" }, + "request": { + "bodyTemplate": { + "query": "${query}", + "max_results": "${maxResults}", + "search_depth": "basic" + } + }, + "response": { + "resultsPath": "results", + "item": { "titlePath": "title", "urlPath": "url", "snippetPath": "content" } + }, + "timeoutMs": 10000 +} diff --git a/catalog/v2/retirements.json b/catalog/v2/retirements.json new file mode 100644 index 0000000..083baff --- /dev/null +++ b/catalog/v2/retirements.json @@ -0,0 +1,24 @@ +{ + "description": "本产品已下架的受管预置接入。规则限定Provider身份与原始端点；不证明厂家根域停服，不删除用户凭据。", + "rules": [ + { + "id": "config-center-lingyiwanwu-access-retirement", + "kind": "product-access-retired", + "scope": "managed-provider", + "providerId": "provider-lingyiwanwu-001", + "origin": { + "exactHost": "api.lingyiwanwu.com", + "pathPrefix": "/v1" + }, + "effectiveAt": "2026-08-28T06:52:05Z", + "reason": "config-center #84 已移除零一万物预置接入；存量受管配置不再作为官方可调用接入。", + "evidence": { + "kind": "product-repository", + "url": "https://github.com/desirecore/config-center/commit/a998e18e6c4bdcc161f0baf82cbf534cc9d53627", + "checkedAt": "2026-10-03" + }, + "preserveCredentials": true, + "vendorDomainShutdownConfirmed": false + } + ] +} diff --git a/compute/coding-plans/_index.json b/compute/coding-plans/_index.json new file mode 100644 index 0000000..3c56407 --- /dev/null +++ b/compute/coding-plans/_index.json @@ -0,0 +1,16 @@ +{ + "description": "Coding Plan / Token Plan 预设加载顺序索引", + "order": [ + "zhipu-coding", + "dashscope-coding", + "dashscope-token-plan", + "volcengine-coding", + "tencent-token", + "minimax-coding", + "moonshot-coding", + "baidu-coding", + "infini-coding", + "moorethread-coding", + "kwai-coding" + ] +} diff --git a/compute/coding-plans/baidu-coding.json b/compute/coding-plans/baidu-coding.json new file mode 100644 index 0000000..593a162 --- /dev/null +++ b/compute/coding-plans/baidu-coding.json @@ -0,0 +1,178 @@ +{ + "id": "provider-baidu-coding-001", + "provider": "baidu", + "brandGroup": "baidu", + "label": "百度千帆 Coding Plan", + "baseUrl": "https://qianfan.baidubce.com/v2/coding", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://console.bce.baidu.com/qianfan/codingplan" + } + }, + "models": [ + { + "modelName": "qianfan-code-latest", + "displayName": "Auto (自动路由)", + "serviceType": [ + "chat" + ], + "description": "千帆 Coding Plan 自动路由模型，后端自动选择最优模型", + "capabilities": [ + "chat", + "reasoning", + "code" + ], + "source": "preset" + }, + { + "modelName": "deepseek-v3.2", + "displayName": "DeepSeek V3.2", + "serviceType": [ + "chat" + ], + "description": "DeepSeek V3.2（千帆 Coding Plan 渠道），倍率 ×1", + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "deepseek-v4-flash", + "displayName": "DeepSeek V4 Flash", + "serviceType": [ + "chat" + ], + "description": "DeepSeek V4 Flash（千帆 Coding Plan 渠道），倍率 ×1", + "capabilities": [ + "chat", + "reasoning", + "code", + "fast" + ], + "source": "preset" + }, + { + "modelName": "ernie-4.5-turbo-20260402", + "displayName": "文心 ERNIE 4.5 Turbo", + "serviceType": [ + "chat" + ], + "description": "百度 ERNIE 4.5 Turbo（千帆 Coding Plan 渠道）；套餐当前不支持图像理解", + "capabilities": [ + "chat", + "code", + "fast" + ], + "source": "preset" + }, + { + "modelName": "glm-5", + "displayName": "GLM-5", + "serviceType": [ + "chat" + ], + "description": "智谱 GLM-5（千帆 Coding Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "glm-5.1", + "displayName": "GLM-5.1", + "serviceType": [ + "chat" + ], + "description": "智谱 GLM-5.1（千帆 Coding Plan 渠道），高峰抵扣 ×4 / 低峰 ×3；高峰时段随流量动态调整", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.5", + "displayName": "Kimi K2.5", + "serviceType": [ + "chat" + ], + "description": "Kimi K2.5（千帆 Coding Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent", + "long_context", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.6", + "displayName": "Kimi K2.6", + "serviceType": [ + "chat" + ], + "description": "Kimi K2.6（千帆 Coding Plan 渠道），倍率 ×5", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "minimax-m2.5", + "displayName": "MiniMax M2.5", + "serviceType": [ + "chat" + ], + "description": "MiniMax M2.5（千帆 Coding Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "minimax-m2.7", + "displayName": "MiniMax M2.7", + "serviceType": [ + "chat" + ], + "description": "MiniMax M2.7（千帆 Coding Plan 渠道），倍率 ×2", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/dashscope-coding.json b/compute/coding-plans/dashscope-coding.json new file mode 100644 index 0000000..5226e96 --- /dev/null +++ b/compute/coding-plans/dashscope-coding.json @@ -0,0 +1,210 @@ +{ + "id": "provider-dashscope-coding-001", + "provider": "dashscope", + "brandGroup": "dashscope", + "label": "百炼 Coding Plan", + "baseUrl": "https://coding.dashscope.aliyuncs.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://bailian.console.aliyun.com/" + } + }, + "models": [ + { + "modelName": "qwen3.7-plus", + "displayName": "Qwen3.7 Plus (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，通义千问 3.7 Plus，多模态智能体模型，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "qwen3.6-plus", + "displayName": "Qwen3.6 Plus (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，通义千问 3.6 Plus，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "qwen3.5-plus", + "displayName": "Qwen3.5 Plus (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，通义千问 3.5 Plus，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "qwen3-coder-plus", + "displayName": "Qwen3 Coder Plus (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，Qwen3 编程增强版，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "qwen3-coder-next", + "displayName": "Qwen3 Coder Next (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，Qwen3 编程下一代实验模型，262K 上下文", + "contextWindow": 262144, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "qwen3-max-2026-01-23", + "displayName": "Qwen3 Max (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，Qwen3 Max 旗舰模型，262K 上下文", + "contextWindow": 262144, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context", + "vision" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.5", + "displayName": "Kimi K2.5 (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，月之暗面 Kimi K2.5，262K 上下文", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "vision", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "glm-5", + "displayName": "GLM-5 (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，智谱 GLM-5，200K 上下文", + "contextWindow": 202752, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent" + ], + "source": "preset" + }, + { + "modelName": "glm-4.7", + "displayName": "GLM-4.7 (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，智谱 GLM-4.7，200K 上下文", + "contextWindow": 202752, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "MiniMax-M2.5", + "displayName": "MiniMax M2.5 (Coding)", + "serviceType": [ + "chat" + ], + "description": "百炼 Coding Plan，MiniMax M2.5，196K 上下文", + "contextWindow": 196608, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "long_context" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/dashscope-token-plan.json b/compute/coding-plans/dashscope-token-plan.json new file mode 100644 index 0000000..a189baf --- /dev/null +++ b/compute/coding-plans/dashscope-token-plan.json @@ -0,0 +1,549 @@ +{ + "id": "provider-dashscope-token-plan-001", + "provider": "dashscope", + "brandGroup": "dashscope", + "label": "百炼 Token Plan", + "baseUrl": "https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "token-plan", + "services": [ + "chat", + "reasoning" + ], + "tombstones": [ + "qwen3.8-max-preview", + "qwen3.5-plus", + "qwen3-coder-plus" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://bailian.console.aliyun.com/" + } + }, + "models": [ + { + "modelName": "deepseek-v4.1-flash", + "displayName": "DeepSeek V4.1 Flash", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "extra": { + "supportsThinking": true, + "thinkingDefault": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "description": "DeepSeek V4.1 Flash，支持文本与图像输入，100 万上下文", + "source": "preset" + }, + { + "modelName": "wan2.7-image-pro", + "displayName": "通义万相 2.7 Pro", + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相最新旗舰文生图模型，支持 4K 分辨率", + "source": "preset", + "extra": {} + }, + { + "modelName": "wan2.7-image", + "displayName": "通义万相 2.7", + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相最新标准文生图模型，高画质", + "source": "preset", + "extra": {} + }, + { + "modelName": "qwen-image-3.0-pro", + "displayName": "qwen-image-3.0-pro", + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0-pro", + "source": "preset" + }, + { + "modelName": "happyhorse-1.1-t2v", + "displayName": "HappyHorse 1.1 T2V", + "capabilities": [ + "video_generation", + "text_to_video", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 文生视频，进一步提升文本词义理解、镜头调度与动态生成表现", + "source": "preset", + "extra": {} + }, + { + "modelName": "happyhorse-1.1-r2v", + "displayName": "HappyHorse 1.1 R2V", + "capabilities": [ + "video_generation", + "reference_to_video", + "multi_image", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 参考生视频，支持最多9张图片参考，更加稳定的主体与场景参考", + "source": "preset", + "extra": {} + }, + { + "modelName": "happyhorse-1.1-i2v", + "displayName": "HappyHorse 1.1 I2V", + "capabilities": [ + "video_generation", + "image_to_video", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 图生视频，进一步提升画面质感、动态表现与跨片段一致性", + "source": "preset", + "extra": {} + }, + { + "modelName": "glm-5.3", + "displayName": "GLM-5.3", + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "extra": { + "thinkingDefault": true, + "thinkingOnly": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "source": "preset" + }, + { + "modelName": "glm-5.2", + "displayName": "GLM-5.2", + "contextWindow": 1048576, + "maxOutputTokens": 32768, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "智谱 GLM-5.2 大推理模型，100 万上下文", + "source": "preset", + "extra": { + "reasoning": { + "supportedEfforts": [ + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultEffort": "high" + } + } + }, + { + "modelName": "deepseek-v4-pro-0813", + "displayName": "DeepSeek V4 Pro 0813", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "extra": { + "supportsThinking": true, + "thinkingDefault": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "description": "DeepSeek V4 Pro 0813 版本，1.6T MoE 架构，100 万上下文", + "source": "preset" + }, + { + "modelName": "deepseek-v4-pro", + "displayName": "DeepSeek V4 Pro", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "extra": { + "supportsThinking": true, + "thinkingDefault": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "description": "DeepSeek V4 Pro，1.6T MoE 架构，100 万上下文", + "source": "preset" + }, + { + "modelName": "deepseek-v4-flash-0731", + "displayName": "DeepSeek V4 Flash 0731", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "extra": { + "supportsThinking": true, + "thinkingDefault": true, + "reasoning": { + "supportedEfforts": [ + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "description": "DeepSeek V4 Flash 0731 固定版本，100 万上下文", + "source": "preset" + }, + { + "modelName": "qwen3.8-max", + "displayName": "Qwen3.8 Max (Token Plan)", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "百炼 Token Plan，通义千问 3.8 Max 正式版，支持文本、图像和视频输入", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "defaultTemperature": 0.6, + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + }, + "supportsThinking": true, + "thinkingDefault": true, + "thinkingMaxTokens": 262144, + "preserveThinkingDefault": true, + "supportsParallelToolCalls": false + }, + "source": "preset" + }, + { + "modelName": "qwen3.8-flash", + "displayName": "Qwen3.8 Flash (Token Plan)", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "百炼 Token Plan，通义千问 3.8 Flash 正式版，高速多模态推理", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "tool_use", + "agent", + "long_context", + "fast" + ], + "defaultTemperature": 0.6, + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + }, + "supportsThinking": true, + "thinkingDefault": true, + "thinkingMaxTokens": 262144, + "preserveThinkingDefault": true, + "supportsParallelToolCalls": false + }, + "source": "preset" + }, + { + "modelName": "qwen3.7-max", + "displayName": "Qwen3.7 Max (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，通义千问 3.7 Max，1M 上下文，面向复杂推理与长程 Agent 任务", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "qwen3.7-plus", + "displayName": "Qwen3.7 Plus (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，通义千问 3.7 Plus，多模态智能体模型，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "qwen3.6-plus", + "displayName": "Qwen3.6 Plus (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，通义千问 3.6 Plus，1M 上下文，支持图像理解", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "qwen3.6-flash", + "displayName": "Qwen3.6 Flash (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，通义千问 3.6 Flash，高性价比多模态模型，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context", + "fast" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.5", + "displayName": "Kimi K2.5 (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，Kimi K2.5，支持图像理解", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "glm-5", + "displayName": "GLM-5 (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，智谱 GLM-5，200K 上下文", + "contextWindow": 202752, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent" + ], + "source": "preset" + }, + { + "modelName": "MiniMax-M2.5", + "displayName": "MiniMax M2.5 (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，MiniMax M2.5，196K 上下文", + "contextWindow": 196608, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "glm-4.7", + "displayName": "GLM-4.7 (Token Plan)", + "serviceType": [ + "chat" + ], + "description": "百炼 Token Plan，智谱 GLM-4.7，200K 上下文", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/infini-coding.json b/compute/coding-plans/infini-coding.json new file mode 100644 index 0000000..7719329 --- /dev/null +++ b/compute/coding-plans/infini-coding.json @@ -0,0 +1,43 @@ +{ + "id": "provider-infini-coding-001", + "provider": "infini", + "brandGroup": "infini", + "label": "无问芯穹 Token (Coding) Plan", + "baseUrl": "https://cloud.infini-ai.com/maas/coding/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://cloud.infini-ai.com/console" + } + }, + "models": [ + { + "modelName": "deepseek-v3", + "displayName": "DeepSeek V3 (芯穹 Coding)", + "serviceType": [ + "chat" + ], + "description": "无问芯穹 Coding Plan，DeepSeek V3 编程优化", + "contextWindow": 128000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/kwai-coding.json b/compute/coding-plans/kwai-coding.json new file mode 100644 index 0000000..38f4fd2 --- /dev/null +++ b/compute/coding-plans/kwai-coding.json @@ -0,0 +1,38 @@ +{ + "id": "provider-kwai-coding-001", + "provider": "kwai", + "brandGroup": "kwai", + "label": "快手 KwaiKAT Token (Coding) Plan", + "baseUrl": "https://wanqing.streamlakeapi.com/api/gateway/coding/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://console.streamlake.ai/wanqing/" + } + }, + "models": [ + { + "modelName": "kwai-coder", + "displayName": "KwaiKAT Coder (Coding)", + "serviceType": [ + "chat" + ], + "description": "快手 KwaiKAT Coding Plan，编程辅助模型", + "capabilities": [ + "chat", + "code" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/minimax-coding.json b/compute/coding-plans/minimax-coding.json new file mode 100644 index 0000000..8257ba4 --- /dev/null +++ b/compute/coding-plans/minimax-coding.json @@ -0,0 +1,226 @@ +{ + "id": "provider-minimax-coding-001", + "provider": "minimax", + "brandGroup": "minimax", + "label": "MiniMax Token (Coding) Plan", + "baseUrl": "https://api.minimaxi.com/anthropic", + "apiFormat": "anthropic-messages", + "mediaBaseUrl": "https://api.minimaxi.com/v1", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "fast", + "reasoning", + "image_gen", + "tts", + "video_gen", + "music_gen" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "rest-api", + "endpoint": "https://api.minimaxi.com/v1/token_plan/remains", + "consoleUrl": "https://platform.minimaxi.com/" + } + }, + "models": [ + { + "modelName": "MiniMax-M3.1-Flash-Preview", + "displayName": "MiniMax M3.1 Flash Preview", + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "MiniMax M3.1 Flash Preview，目前仅通过 M Plan / MiniMax Code 提供", + "contextWindow": 1000000, + "extra": { + "adaptiveThinking": true, + "thinkingOnly": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "max" + } + }, + "source": "preset" + }, + { + "modelName": "MiniMax-M3", + "displayName": "MiniMax M3 (Token Plan)", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "MiniMax M3，Token Plan 核心文本模型，1M 上下文，面向 Coding 与 Agent", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "vision", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "MiniMax-M2.7", + "displayName": "MiniMax M2.7 (Token Plan)", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "MiniMax 新一代旗舰模型，Token Plan 核心文本模型，5 小时滚动窗口计费。Starter 600次/5h，Plus 1500次/5h，Max 4500次/5h", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "MiniMax-M2.7-highspeed", + "displayName": "MiniMax M2.7 高速版 (Token Plan)", + "serviceType": [ + "fast" + ], + "description": "MiniMax M2.7 极速版，极速套餐专用。Plus-极速 1500次/5h，Max-极速 4500次/5h，Ultra-极速 30000次/5h", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "source": "preset" + }, + { + "modelName": "speech-2.8-hd", + "displayName": "Speech 2.8 HD (Token Plan)", + "serviceType": [ + "tts" + ], + "description": "MiniMax 最新高清语音合成，Token Plan Plus 及以上可用。Plus 4000字符/天，Max 11000字符/天", + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone" + ], + "source": "preset" + }, + { + "modelName": "speech-2.8-turbo", + "displayName": "Speech 2.8 Turbo (Token Plan)", + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成，Token Plan Plus 及以上可用", + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast" + ], + "source": "preset" + }, + { + "modelName": "image-01", + "displayName": "Image 01 (Token Plan)", + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图/图生图，Token Plan Plus 及以上可用。Plus 50张/天，Max 120张/天", + "capabilities": [ + "image_generation", + "subject_reference" + ], + "source": "preset" + }, + { + "modelName": "image-01-live", + "displayName": "Image 01 Live (Token Plan)", + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 画风控制文生图，Token Plan Plus 及以上可用", + "capabilities": [ + "image_generation", + "style_control", + "subject_reference" + ], + "source": "preset" + }, + { + "modelName": "MiniMax-Hailuo-2.3", + "displayName": "海螺视频 2.3 (Token Plan)", + "serviceType": [ + "video_gen" + ], + "description": "海螺视频 2.3，文/图生视频，Token Plan Max 及以上可用。Max 2个/天", + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "camera_control" + ], + "source": "preset" + }, + { + "modelName": "MiniMax-Hailuo-2.3-Fast", + "displayName": "海螺视频 2.3 快速版 (Token Plan)", + "serviceType": [ + "video_gen" + ], + "description": "海螺视频快速版，Token Plan Max 及以上可用。Max 2个/天", + "capabilities": [ + "video_generation", + "image_to_video", + "fast" + ], + "source": "preset" + }, + { + "modelName": "music-2.6", + "displayName": "Music 2.6 (Token Plan)", + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成，所有 Token Plan 套餐免费使用（100 首/天，≤5分钟/首）", + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/moonshot-coding.json b/compute/coding-plans/moonshot-coding.json new file mode 100644 index 0000000..9c5119f --- /dev/null +++ b/compute/coding-plans/moonshot-coding.json @@ -0,0 +1,33 @@ +{ + "id": "provider-moonshot-coding-001", + "provider": "moonshot", + "brandGroup": "moonshot", + "label": "Kimi Code", + "baseUrl": "https://api.kimi.com/coding/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": ["chat", "reasoning"], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://platform.moonshot.cn/" + } + }, + "models": [ + { + "modelName": "kimi-for-coding", + "displayName": "Kimi For Coding (Auto)", + "serviceType": ["chat", "reasoning"], + "description": "Kimi Code 稳定路由 ID，自动映射到 Kimi Code 当前模型，262K 上下文", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": ["chat", "reasoning", "code", "tool_use", "agent", "long_context", "vision"], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/moorethread-coding.json b/compute/coding-plans/moorethread-coding.json new file mode 100644 index 0000000..5e191bb --- /dev/null +++ b/compute/coding-plans/moorethread-coding.json @@ -0,0 +1,38 @@ +{ + "id": "provider-moorethread-coding-001", + "provider": "moorethread", + "brandGroup": "moorethread", + "label": "摩尔线程 Token (Coding) Plan", + "baseUrl": "https://api.mthreads.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://www.mthreads.com/console" + } + }, + "models": [ + { + "modelName": "mt-coder", + "displayName": "MT Coder (Coding)", + "serviceType": [ + "chat" + ], + "description": "摩尔线程 Coding Plan，国产 GPU 推理", + "capabilities": [ + "chat", + "code" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/tencent-token.json b/compute/coding-plans/tencent-token.json new file mode 100644 index 0000000..06263ad --- /dev/null +++ b/compute/coding-plans/tencent-token.json @@ -0,0 +1,180 @@ +{ + "id": "provider-tencent-token-001", + "provider": "tencent", + "brandGroup": "tencent", + "label": "腾讯云 Token Plan", + "baseUrl": "https://api.lkeap.cloud.tencent.com/plan/v3", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "token-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://console.cloud.tencent.com/hunyuan" + } + }, + "models": [ + { + "modelName": "tc-code-latest", + "displayName": "Auto (自动路由)", + "serviceType": [ + "chat" + ], + "description": "腾讯 Token Plan 自动路由模型，后端自动选择最优模型", + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "hunyuan-2.0-instruct", + "displayName": "混元 2.0 Instruct", + "serviceType": [ + "chat" + ], + "description": "腾讯混元 2.0 指令模型（Token Plan）", + "capabilities": [ + "chat", + "reasoning", + "code", + "long_context", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "hunyuan-2.0-thinking", + "displayName": "混元 2.0 Think", + "serviceType": [ + "chat" + ], + "description": "腾讯混元 2.0 推理模型（Token Plan）", + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "hunyuan-t1", + "displayName": "混元 T1", + "serviceType": [ + "chat" + ], + "description": "腾讯混元 T1 深度推理模型（Token Plan）", + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking" + ], + "source": "preset" + }, + { + "modelName": "hunyuan-turbos", + "displayName": "混元 Turbo S", + "serviceType": [ + "chat" + ], + "description": "腾讯混元 Turbo S 高速模型（Token Plan）", + "capabilities": [ + "chat", + "reasoning", + "code", + "fast" + ], + "source": "preset" + }, + { + "modelName": "glm-5.1", + "displayName": "GLM-5.1", + "serviceType": [ + "chat" + ], + "description": "智谱 GLM-5.1（腾讯 Token Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent" + ], + "source": "preset" + }, + { + "modelName": "glm-5", + "displayName": "GLM-5", + "serviceType": [ + "chat" + ], + "description": "智谱 GLM-5（腾讯 Token Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent" + ], + "source": "preset" + }, + { + "modelName": "minimax-m2.7", + "displayName": "MiniMax M2.7", + "serviceType": [ + "chat" + ], + "description": "MiniMax M2.7（腾讯 Token Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "minimax-m2.5", + "displayName": "MiniMax M2.5", + "serviceType": [ + "chat" + ], + "description": "MiniMax M2.5（腾讯 Token Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.5", + "displayName": "Kimi K2.5", + "serviceType": [ + "chat" + ], + "description": "Kimi K2.5（腾讯 Token Plan 渠道）", + "capabilities": [ + "chat", + "reasoning", + "code", + "agent", + "long_context" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/volcengine-coding.json b/compute/coding-plans/volcengine-coding.json new file mode 100644 index 0000000..a87b979 --- /dev/null +++ b/compute/coding-plans/volcengine-coding.json @@ -0,0 +1,259 @@ +{ + "id": "provider-volcengine-coding-001", + "provider": "volcengine", + "brandGroup": "volcengine", + "label": "火山引擎 Token (Coding) Plan", + "baseUrl": "https://ark.cn-beijing.volces.com/api/coding/v3", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "response-header", + "headerKeys": { + "remaining": "x-ratelimit-remaining", + "limit": "x-ratelimit-limit", + "reset": "x-ratelimit-reset" + }, + "consoleUrl": "https://console.volcengine.com/ark" + }, + "modelIdOverride": "ark-code-latest" + }, + "models": [ + { + "modelName": "doubao-seed-2.1-pro", + "displayName": "豆包 Seed 2.1 Pro", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan，豆包 Seed 2.1 Pro，生产级 Coding 与 Agent 模型", + "contextWindow": 256000, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "doubao-seed-2.1-turbo", + "displayName": "豆包 Seed 2.1 Turbo", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan，豆包 Seed 2.1 Turbo，高吞吐 Coding 与 Agent 模型", + "contextWindow": 256000, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "fast", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "ark-code-latest", + "displayName": "Ark Code Latest (Auto)", + "serviceType": [ + "chat" + ], + "description": "自动路由最优编程模型（Doubao Seed Code / GLM-4.7 / DeepSeek-V3.2 / Kimi-K2.5 等）", + "capabilities": [ + "code", + "reasoning", + "chat", + "tool_use" + ], + "source": "preset" + }, + { + "modelName": "doubao-seed-2.0-code", + "displayName": "豆包 Seed 2.0 Code", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的豆包 Seed 2.0 编程模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "doubao-seed-2.0-pro", + "displayName": "豆包 Seed 2.0 Pro", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的豆包 Seed 2.0 旗舰模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "long_context", + "vision" + ], + "source": "preset" + }, + { + "modelName": "doubao-seed-2.0-lite", + "displayName": "豆包 Seed 2.0 Lite", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的豆包 Seed 2.0 均衡模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "long_context", + "vision" + ], + "source": "preset" + }, + { + "modelName": "doubao-seed-code", + "displayName": "豆包 Seed Code", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的豆包编程模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "glm-4.7", + "displayName": "GLM-4.7", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的智谱 GLM-4.7 模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "deepseek-v3.2", + "displayName": "DeepSeek V3.2", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的 DeepSeek V3.2 模型", + "capabilities": [ + "code", + "reasoning", + "chat" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.5", + "displayName": "Kimi K2.5", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的 Kimi K2.5 模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "agent", + "long_context" + ], + "source": "preset" + }, + { + "modelName": "glm-5.1", + "displayName": "GLM-5.1", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的智谱 GLM-5.1 旗舰模型（2026.4 新增）", + "capabilities": [ + "code", + "reasoning", + "chat", + "long_context", + "agent" + ], + "source": "preset" + }, + { + "modelName": "kimi-k2.6", + "displayName": "Kimi K2.6", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的 Kimi K2.6 模型（2026.4 新增）", + "capabilities": [ + "code", + "reasoning", + "chat", + "agent", + "long_context", + "vision" + ], + "source": "preset" + }, + { + "modelName": "minimax-m2.5", + "displayName": "MiniMax M2.5", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的 MiniMax M2.5 模型", + "capabilities": [ + "code", + "reasoning", + "chat", + "agent" + ], + "source": "preset" + }, + { + "modelName": "minimax-m2.7", + "displayName": "MiniMax M2.7", + "serviceType": [ + "chat" + ], + "description": "火山方舟 Coding Plan 中的 MiniMax M2.7 模型（2026.4 新增），230B 总参/10B 激活", + "capabilities": [ + "code", + "reasoning", + "chat", + "agent", + "long_context" + ], + "source": "preset" + } + ] +} diff --git a/compute/coding-plans/zhipu-coding.json b/compute/coding-plans/zhipu-coding.json new file mode 100644 index 0000000..9b084e8 --- /dev/null +++ b/compute/coding-plans/zhipu-coding.json @@ -0,0 +1,197 @@ +{ + "id": "provider-zhipu-coding-001", + "provider": "zhipu", + "brandGroup": "zhipu", + "label": "智谱 Token (Coding) Plan", + "baseUrl": "https://api.z.ai/api/anthropic", + "apiFormat": "anthropic-messages", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://open.bigmodel.cn/console/overview" + } + }, + "tombstones": [ + "glm-5.1", + "glm-5", + "glm-4.5-air" + ], + "models": [ + { + "modelName": "glm-5.3", + "displayName": "GLM-5.3", + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "extra": { + "adaptiveThinking": true, + "thinkingDefault": true, + "thinkingOnly": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + }, + "pricingNotes": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。" + }, + "source": "preset" + }, + { + "modelName": "glm-5.3-flash", + "displayName": "GLM-5.3-Flash (Coding)", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "智谱 GLM-5.3-Flash Coding Plan 版本，原生多模态高效模型，1M 上下文，128K 最大输出，强制开启思考", + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "image_understanding", + "video_understanding", + "long_context", + "tool_use", + "agent", + "multilingual", + "fast" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "adaptiveThinking": true, + "supportsThinking": true, + "thinkingDefault": true, + "thinkingOnly": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "source": "preset" + }, + { + "modelName": "glm-5.2", + "apiModelId": "glm-5.2[1m]", + "displayName": "GLM-5.2 (Coding)", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-5.2 Coding Plan 版本，官方最新可用模型，1M 上下文，128K 最大输出", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "math", + "multilingual" + ], + "extra": { + "codingPlanLongContextModelId": "glm-5.2[1m]", + "reasoning": { + "supportedEfforts": [ + "max", + "xhigh", + "high", + "medium", + "low", + "minimal", + "none" + ], + "defaultEffort": "max" + } + }, + "source": "preset" + }, + { + "modelName": "glm-5-turbo", + "displayName": "GLM-5-Turbo (Coding)", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-5-Turbo Coding Plan 版本，官方可用高阶模型，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "math", + "multilingual", + "fast" + ], + "source": "preset" + }, + { + "modelName": "glm-4.7", + "displayName": "GLM-4.7 (Coding)", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-4.7 Coding Plan 版本，官方可用模型，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "multilingual" + ], + "source": "preset" + } + ] +} diff --git a/compute/model-specs/_index.json b/compute/model-specs/_index.json new file mode 100644 index 0000000..9fe2b20 --- /dev/null +++ b/compute/model-specs/_index.json @@ -0,0 +1,63 @@ +{ + "description": "模型规格库加载顺序索引，以及智能路由三档的同域定义。模型级路由策略仅写入各 ModelSpec 的 routing 字段，不另建 smart-routing 目录。 数据批次 130：GLM-5.3 系列在 anthropic-messages 协议上按智谱官网以 thinking.type 与 output_config.effort 控制思考强度（protocolProfiles 声明 adaptive 形态），路由策略不变；此前批次 129 按官网补对话规格的 tool_use 标签，批次 126 为官方字段复核与临时、历史规格边界。 精确身份治理：隔离Nano/实时/固定变体，identity-only不继承family参数；发布批次由manifest统一标记。", + "order": [ + "anthropic", + "openai", + "google", + "deepseek", + "qwen", + "wan", + "moonshot", + "zhipu", + "minimax", + "xai", + "mistral", + "xiaomi", + "baichuan", + "baidu", + "cohere", + "kling", + "perplexity", + "stability", + "tencent", + "volcengine", + "xunfei", + "stealth", + "happyhorse" + ], + "routingTiers": [ + { + "id": "flagship", + "label": "旗舰", + "labelEn": "Flagship", + "description": "最高能力，适合复杂推理、关键交付和长程智能体任务。", + "fallbackOrder": [ + "flagship", + "balanced", + "lightweight" + ] + }, + { + "id": "balanced", + "label": "均衡", + "labelEn": "Balanced", + "description": "质量、速度与成本平衡，适合大多数日常任务。", + "fallbackOrder": [ + "balanced", + "flagship", + "lightweight" + ] + }, + { + "id": "lightweight", + "label": "轻量", + "labelEn": "Lightweight", + "description": "优先低延迟与低成本，适合明确、短链路和批量任务。", + "fallbackOrder": [ + "lightweight", + "balanced", + "flagship" + ] + } + ] +} diff --git a/compute/model-specs/anthropic.json b/compute/model-specs/anthropic.json new file mode 100644 index 0000000..27e5386 --- /dev/null +++ b/compute/model-specs/anthropic.json @@ -0,0 +1,419 @@ +{ + "description": "Anthropic Claude 系列模型规格。参数来源：config-center compute/providers/anthropic.json。", + "specs": [ + { + "id": "claude-opus-5-5", + "displayName": "Claude Opus 5.5", + "family": "claude-opus-5-5", + "match": { + "exact": [ + "claude-opus-5-5", + "anthropic/claude-opus-5.5" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Opus 5.5，始终开启自适应思考，默认 medium effort", + "extra": { + "adaptiveThinking": true, + "thinkingOnly": true, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + }, + "releasedAt": "2026-09-22" + }, + "routing": { + "tier": "flagship", + "routingPriority": 22, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "claude-fable-5-1", + "displayName": "Claude Fable 5.1", + "family": "claude-fable", + "match": { + "exact": [ + "claude-fable-5-1" + ], + "patterns": [ + "claude-fable-5-1*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Fable 5.1，面向高难推理和长程智能体任务，始终开启自适应思考", + "extra": { + "adaptiveThinking": true, + "thinkingOnly": true, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "claude-opus-5", + "displayName": "Claude Opus 5", + "family": "claude-opus-5", + "match": { + "exact": [ + "claude-opus-5" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Opus 5，Anthropic Opus 系列当前旗舰；思考默认开启，effort 支持到 max", + "extra": { + "adaptiveThinking": true, + "samplingParametersDeprecated": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 25, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "xhigh" + } + } + }, + { + "id": "claude-opus-4-8", + "displayName": "Claude Opus 4.8", + "family": "claude-opus-4-8", + "match": { + "patterns": [ + "claude-opus-4-8*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Opus 4.8，Anthropic Opus 系列最新旗舰", + "extra": { + "adaptiveThinking": true, + "samplingParametersDeprecated": true + } + } + }, + { + "id": "claude-opus-4-7", + "displayName": "Claude Opus 4.7", + "family": "claude-opus-4-7", + "match": { + "patterns": [ + "claude-opus-4-7*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "Anthropic 当前最强通用模型，适合复杂推理和智能体编码任务", + "extra": { + "adaptiveThinking": true, + "samplingParametersDeprecated": true + } + } + }, + { + "id": "claude-opus", + "family": "claude-opus", + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + } + }, + { + "id": "claude-sonnet-5-5", + "displayName": "Claude Sonnet 5.5", + "family": "claude-sonnet-5-5", + "match": { + "exact": [ + "claude-sonnet-5-5", + "anthropic/claude-sonnet-5.5" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Sonnet 5.5，速度与智能平衡，默认 high effort，自适应思考", + "extra": { + "adaptiveThinking": true, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + }, + "releasedAt": "2026-09-28" + }, + "routing": { + "tier": "balanced", + "routingPriority": 18, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "claude-sonnet-5", + "displayName": "Claude Sonnet 5", + "family": "claude-sonnet-5", + "match": { + "patterns": [ + "claude-sonnet-5*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Sonnet 5，速度与智能平衡；思考默认开启，effort 支持到 max", + "extra": { + "adaptiveThinking": true, + "samplingParametersDeprecated": true + } + }, + "routing": { + "tier": "balanced", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "xhigh" + } + } + }, + { + "id": "claude-sonnet-4-6", + "displayName": "Claude Sonnet 4.6", + "family": "claude-sonnet-4-6", + "match": { + "patterns": [ + "claude-sonnet-4-6*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 64000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use" + ], + "serviceType": [ + "chat", + "computer_use" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Anthropic 高智能高速度模型，适合编码、工具使用和智能体任务" + } + }, + { + "id": "claude-sonnet-4-5", + "displayName": "Claude Sonnet 4.5", + "family": "claude-sonnet-4-5", + "match": { + "patterns": [ + "claude-sonnet-4-5*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 64000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Anthropic 主力模型，长文本与代码任务表现优异" + } + }, + { + "id": "claude-sonnet", + "family": "claude-sonnet", + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + } + } + ] +} diff --git a/compute/model-specs/baichuan.json b/compute/model-specs/baichuan.json new file mode 100644 index 0000000..d88a8ec --- /dev/null +++ b/compute/model-specs/baichuan.json @@ -0,0 +1,111 @@ +{ + "description": "百川智能 Baichuan 系列模型规格。参数来源：config-center compute/providers/baichuan.json。M3/M2 系列为医疗增强大模型，235B 参数，支持多模态。", + "specs": [ + { + "id": "Baichuan-M3-Plus", + "displayName": "百川 M3-Plus", + "family": "baichuan-m3", + "match": { + "exact": [ + "Baichuan-M3-Plus" + ] + }, + "spec": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "long_context", + "medical" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M3-Plus医疗增强模型，235B参数，幻觉率仅2.6%" + } + }, + { + "id": "Baichuan-M3", + "displayName": "百川 M3", + "family": "baichuan-m3", + "match": { + "exact": [ + "Baichuan-M3" + ] + }, + "spec": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "medical" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M3新一代开源医疗增强模型，235B参数" + } + }, + { + "id": "Baichuan-M2-Plus", + "displayName": "百川 M2-Plus", + "family": "baichuan-m2", + "match": { + "exact": [ + "Baichuan-M2-Plus" + ] + }, + "spec": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M2-Plus旗舰模型，最强推理和创作能力" + } + }, + { + "id": "Baichuan-M2", + "displayName": "百川 M2", + "family": "baichuan-m2", + "match": { + "exact": [ + "Baichuan-M2" + ] + }, + "spec": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M2高性能通用模型" + } + } + ] +} diff --git a/compute/model-specs/baidu.json b/compute/model-specs/baidu.json new file mode 100644 index 0000000..e23b24c --- /dev/null +++ b/compute/model-specs/baidu.json @@ -0,0 +1,146 @@ +{ + "description": "百度文心 ERNIE 系列模型规格。参数来源：config-center compute/providers/baidu.json。ernie-x1.1 为纯推理模型，defaultTemperature 为 null；ernie-5.0-thinking-latest 同时支持 chat 和 reasoning。", + "specs": [ + { + "id": "ernie-5.0-thinking-latest", + "displayName": "文心 ERNIE 5.0", + "family": "ernie-5", + "match": { + "exact": [ + "ernie-5.0-thinking-latest" + ], + "patterns": [ + "ernie-5.0-thinking*" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "vision", + "multimodal", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "百度最新旗舰模型，支持深度推理、多模态和工具调用" + } + }, + { + "id": "ernie-5.0", + "displayName": "文心 ERNIE 5.0 非思考", + "family": "ernie-5", + "match": { + "exact": [ + "ernie-5.0" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "vision", + "multimodal", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.95, + "description": "百度 ERNIE 5.0 旗舰模型，128K 上下文，最大输出 65536 token" + } + }, + { + "id": "ernie-4.5-turbo-128k", + "displayName": "文心 ERNIE 4.5 Turbo", + "family": "ernie-4.5", + "match": { + "exact": [ + "ernie-4.5-turbo-128k" + ], + "patterns": [ + "ernie-4.5-turbo-128k*" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 12288, + "capabilities": [ + "chat", + "code", + "vision", + "long_context", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.8, + "description": "百度高性价比长上下文模型，128K 窗口，最大输出 12288 token" + } + }, + { + "id": "ernie-4.5-turbo-20260402", + "displayName": "文心 ERNIE 4.5 Turbo 20260402", + "family": "ernie-4.5", + "match": { + "exact": [ + "ernie-4.5-turbo-20260402" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 12288, + "capabilities": [ + "chat", + "code", + "vision", + "long_context", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.8, + "description": "百度 ERNIE 4.5 Turbo 20260402 快照版，128K 上下文" + } + }, + { + "id": "ernie-x1.1", + "displayName": "文心 ERNIE X1.1", + "family": "ernie-x1", + "match": { + "exact": [ + "ernie-x1.1" + ], + "patterns": [ + "ernie-x1*" + ] + }, + "spec": { + "contextWindow": 65536, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "math", + "code", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "百度 ERNIE X1.1 深度思考推理模型，64K 上下文" + } + } + ] +} diff --git a/compute/model-specs/cohere.json b/compute/model-specs/cohere.json new file mode 100644 index 0000000..f9beef7 --- /dev/null +++ b/compute/model-specs/cohere.json @@ -0,0 +1,297 @@ +{ + "description": "Cohere 系列模型规格。参数来源：config-center compute/providers/cohere.json。包含对话、Embedding 和 Rerank 三类服务。", + "specs": [ + { + "id": "rerank-v4.0-fast", + "displayName": "Rerank 4 Fast", + "family": "rerank-v4.0-fast", + "match": { + "exact": [ + "rerank-v4.0-fast" + ] + }, + "spec": { + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Rerank 4 Fast" + } + }, + { + "id": "embed-v5.0-fast", + "displayName": "Embed 5 Fast", + "family": "embed-v5.0-fast", + "match": { + "exact": [ + "embed-v5.0-fast" + ] + }, + "spec": { + "capabilities": [ + "text_embedding", + "vision", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Embed 5 Fast", + "contextWindow": 128000, + "extra": { + "dimensions": [ + 256, + 512, + 768, + 1024, + 1536, + 2048 + ], + "defaultDimension": 2048 + }, + "releasedAt": "2026-09-30" + } + }, + { + "id": "rerank-v4.0-pro", + "displayName": "Rerank 4 Pro", + "family": "rerank-v4.0-pro", + "match": { + "exact": [ + "rerank-v4.0-pro" + ] + }, + "spec": { + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Rerank 4 Pro" + } + }, + { + "id": "embed-v5.0-pro", + "displayName": "Embed 5 Pro", + "family": "embed-v5.0-pro", + "match": { + "exact": [ + "embed-v5.0-pro" + ] + }, + "spec": { + "capabilities": [ + "text_embedding", + "vision", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Embed 5 Pro", + "contextWindow": 128000, + "extra": { + "dimensions": [ + 256, + 512, + 768, + 1024, + 1536, + 2048 + ], + "defaultDimension": 2048 + }, + "releasedAt": "2026-09-30" + } + }, + { + "id": "command-a-plus-05-2026", + "displayName": "Command A+", + "family": "command-a-plus-05-2026", + "match": { + "exact": [ + "command-a-plus-05-2026" + ] + }, + "spec": { + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "multilingual", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "Command A+", + "contextWindow": 128000, + "maxOutputTokens": 64000, + "supportsReasoning": true + } + }, + { + "id": "north-mini-code-1-0", + "displayName": "North Mini Code", + "family": "north-mini-code-1-0", + "match": { + "exact": [ + "north-mini-code-1-0", + "north-mini-code" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 64000, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "description": "Cohere North Mini Code 1.0，原厂API精确ID为north-mini-code-1-0；north-mini-code仅保留本仓历史配置匹配，不声明原厂接受该简称" + } + }, + { + "id": "north-small-translate-1-0", + "displayName": "North Small Translate 1.0", + "family": "north-small-translate-1-0", + "match": { + "exact": [ + "north-small-translate-1-0" + ] + }, + "spec": { + "serviceType": [ + "chat" + ], + "capabilities": [ + "chat", + "multilingual" + ], + "description": "Cohere原厂机器翻译模型，支持50多种语言；官网窗口与输出为16k简写，本轮不猜测精确整数。只登记共享规格，原生接入与账号权限需独立验证" + } + }, + { + "id": "command-a-03-2025", + "displayName": "Command A", + "family": "command-a-03-2025", + "match": { + "exact": [ + "command-a-03-2025", + "cohere/command-a" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 8000, + "capabilities": [ + "chat", + "code", + "tool_use", + "rag", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "Cohere Command A 文本模型，256K 上下文，支持 RAG 和工具调用" + } + }, + { + "id": "command-r7b-12-2024", + "displayName": "Command R7B", + "family": "command-r7b", + "match": { + "exact": [ + "command-r7b-12-2024" + ], + "patterns": [ + "command-r7b*" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 4000, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "rag", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 0.3, + "description": "Cohere 小型高速对话模型，适合高吞吐、低延迟场景" + } + }, + { + "id": "embed-v4.0", + "displayName": "Embed V4", + "family": "embed-v4", + "match": { + "exact": [ + "embed-v4.0" + ], + "patterns": [ + "embed-v4*" + ] + }, + "spec": { + "contextWindow": 128000, + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Cohere 最新 Embedding 模型，128K 上下文，多语言支持", + "extra": { + "dimensions": [ + 256, + 512, + 1024, + 1536 + ], + "defaultDimension": 1536 + } + } + }, + { + "id": "rerank-v3.5", + "displayName": "Rerank V3.5", + "family": "rerank-v3", + "match": { + "exact": [ + "rerank-v3.5" + ], + "patterns": [ + "rerank-v3*" + ] + }, + "spec": { + "contextWindow": 4096, + "capabilities": [ + "rerank", + "semantic_reranking" + ], + "serviceType": [ + "rerank" + ], + "description": "Cohere 语义重排序模型，用于检索结果精排" + } + } + ] +} diff --git a/compute/model-specs/deepseek.json b/compute/model-specs/deepseek.json new file mode 100644 index 0000000..e006210 --- /dev/null +++ b/compute/model-specs/deepseek.json @@ -0,0 +1,354 @@ +{ + "description": "DeepSeek 系列模型规格。参数来源：config-center compute/providers/deepseek.json。", + "specs": [ + { + "id": "deepseek-v4-pro", + "displayName": "DeepSeek V4 Pro", + "family": "deepseek-v4-pro", + "match": { + "exact": [ + "deepseek-v4-pro" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "extra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "description": "DeepSeek V4 Pro，1.6T MoE 架构，100 万上下文" + }, + "routing": { + "tier": "flagship", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": true, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "deepseek-v4-pro-0813", + "displayName": "DeepSeek V4 Pro 0813", + "family": "deepseek-v4-pro-0813", + "match": { + "exact": [ + "deepseek-v4-pro-0813" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "extra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "description": "DeepSeek V4 Pro 0813 版本，1.6T MoE 架构，100 万上下文" + }, + "routing": { + "tier": "flagship", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "deepseek-chat", + "displayName": "DeepSeek Chat", + "family": "deepseek-chat", + "match": { + "patterns": [ + "deepseek-chat*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1 + } + }, + { + "id": "deepseek-reasoner", + "displayName": "DeepSeek Reasoner", + "family": "deepseek-reasoner", + "match": { + "patterns": [ + "deepseek-reasoner*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + } + }, + { + "id": "deepseek-flash", + "displayName": "DeepSeek V4.1 Flash", + "family": "deepseek-flash", + "match": { + "exact": [ + "deepseek-flash", + "deepseek-v4.1-flash", + "deepseek/deepseek-v4.1-flash" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "extra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "description": "DeepSeek V4.1 Flash，支持文本与图像输入，100 万上下文" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "deepseek-v4-flash", + "displayName": "DeepSeek V4 Flash", + "family": "deepseek-v4-flash", + "match": { + "exact": [ + "deepseek-v4-flash" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "extra": { + "supportsThinking": true, + "thinkingDefault": true + } + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "deepseek-v4-flash-0731", + "displayName": "DeepSeek V4 Flash 0731", + "family": "deepseek-v4-flash-0731", + "match": { + "exact": [ + "deepseek-v4-flash-0731" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "extra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "description": "DeepSeek V4 Flash 0731 固定版本，100 万上下文" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "high", + "max" + ], + "defaultMode": "high" + } + } + }, + { + "id": "deepseek-v4-flash-vision-exp", + "displayName": "DeepSeek V4 Flash Vision Exp", + "family": "deepseek-v4-flash-vision-exp", + "match": { + "exact": [ + "deepseek-v4-flash-vision-exp" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "releasedAt": "2026-08-21", + "extra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "description": "DeepSeek V4 Flash 实验性视觉版本，支持文本与图像输入，文本能力与 V4 Flash 一致" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + } + } + ] +} diff --git a/compute/model-specs/google.json b/compute/model-specs/google.json new file mode 100644 index 0000000..06991f9 --- /dev/null +++ b/compute/model-specs/google.json @@ -0,0 +1,535 @@ +{ + "description": "Google Gemini 系列模型规格。参数来源：config-center compute/providers/google.json。", + "specs": [ + { + "id": "gemini-embedding-2", + "displayName": "Gemini Embedding 2", + "family": "gemini-embedding-2", + "match": { + "exact": [ + "gemini-embedding-2" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "vision", + "audio_understanding", + "video_understanding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Gemini Embedding 2，原生多模态嵌入，8192 token 输入，共享文本/图像/音视频向量空间", + "extra": { + "dimensions": [ + 768, + 1536, + 3072 + ], + "defaultDimension": 3072 + } + } + }, + { + "id": "gemini-3.5-transcribe-live", + "displayName": "gemini-3.5-transcribe-live", + "family": "gemini-3.5-transcribe-live", + "match": { + "exact": [ + "gemini-3.5-transcribe-live" + ] + }, + "spec": { + "capabilities": [ + "asr", + "realtime" + ], + "serviceType": [ + "asr" + ], + "description": "gemini-3.5-transcribe-live，官方模型目录；原生音频协议需客户端适配" + } + }, + { + "id": "gemini-3.5-transcribe", + "displayName": "gemini-3.5-transcribe", + "family": "gemini-3.5-transcribe", + "match": { + "exact": [ + "gemini-3.5-transcribe" + ] + }, + "spec": { + "capabilities": [ + "asr" + ], + "serviceType": [ + "asr" + ], + "description": "gemini-3.5-transcribe，官方模型目录；原生音频协议需客户端适配" + } + }, + { + "id": "gemini-3.8-live-extended-thinking", + "displayName": "gemini-3.8-live-extended-thinking", + "family": "gemini-3.8-live-extended-thinking", + "match": { + "exact": [ + "gemini-3.8-live-extended-thinking" + ] + }, + "spec": { + "capabilities": [ + "realtime", + "speech_to_speech", + "reasoning" + ], + "serviceType": [ + "realtime_voice" + ], + "description": "gemini-3.8-live-extended-thinking，官方模型目录；原生音频协议需客户端适配" + } + }, + { + "id": "gemini-3.8-live", + "displayName": "gemini-3.8-live", + "family": "gemini-3.8-live", + "match": { + "exact": [ + "gemini-3.8-live" + ] + }, + "spec": { + "capabilities": [ + "realtime", + "speech_to_speech" + ], + "serviceType": [ + "realtime_voice" + ], + "description": "gemini-3.8-live，官方模型目录；原生音频协议需客户端适配" + } + }, + { + "id": "gemini-3.8-flash-lite-tts", + "displayName": "gemini-3.8-flash-lite-tts", + "family": "gemini-3.8-flash-lite-tts", + "match": { + "exact": [ + "gemini-3.8-flash-lite-tts" + ] + }, + "spec": { + "capabilities": [ + "tts", + "fast" + ], + "serviceType": [ + "tts" + ], + "description": "gemini-3.8-flash-lite-tts，官方模型目录；原生音频协议需客户端适配" + } + }, + { + "id": "gemini-3.8-flash-tts", + "displayName": "gemini-3.8-flash-tts", + "family": "gemini-3.8-flash-tts", + "match": { + "exact": [ + "gemini-3.8-flash-tts" + ] + }, + "spec": { + "capabilities": [ + "tts" + ], + "serviceType": [ + "tts" + ], + "description": "gemini-3.8-flash-tts，官方模型目录；原生音频协议需客户端适配", + "contextWindow": 8192, + "maxOutputTokens": 16384 + } + }, + { + "id": "gemini-3.8-flash", + "displayName": "Gemini 3.8 Flash", + "family": "gemini-3.8-flash", + "match": { + "exact": [ + "gemini-3.8-flash" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Gemini 3.8 Flash，稳定版，支持多模态输入和 1M 上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gemini-3.7-flash", + "displayName": "Gemini 3.7 Flash", + "family": "gemini-3.7-flash", + "match": { + "exact": [ + "gemini-3.7-flash" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Gemini 3.7 Flash，稳定版，支持多模态输入和 1M 上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gemini-3.5-flash-lite", + "displayName": "Gemini 3.5 Flash-Lite", + "family": "gemini-3.5-flash-lite", + "match": { + "exact": [ + "gemini-3.5-flash-lite" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Gemini 3.5 Flash-Lite，稳定版，支持多模态输入和 1M 上下文" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gemini-3.1-flash-lite-image", + "displayName": "Gemini 3.1 Flash Lite Image (Nano Banana 2 Lite)", + "family": "gemini-3.1-flash-lite-image", + "match": { + "exact": [ + "gemini-3.1-flash-lite-image" + ] + }, + "spec": { + "contextWindow": 65536, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "image_generation", + "image_editing", + "tool_use" + ], + "serviceType": [ + "image_gen" + ], + "supportsReasoning": true, + "description": "Nano Banana 2 Lite，Google 面向高吞吐图像生成与编辑的轻量模型", + "releasedAt": "2026-06-30" + } + }, + { + "id": "gemini-3.5-flash", + "displayName": "Gemini 3.5 Flash", + "family": "gemini-3.5", + "match": { + "exact": [ + "gemini-3.5-flash" + ], + "patterns": [ + "gemini-3.5-flash*" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Google Gemini 3.5 Flash，全模态高性价比模型，100 万上下文" + } + }, + { + "id": "gemini-3.1-flash-image", + "displayName": "Gemini 3.1 Flash Image (Nano Banana 2)", + "family": "gemini-3.1", + "match": { + "exact": [ + "gemini-3.1-flash-image" + ], + "patterns": [ + "gemini-3.1-flash-image*" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 32768, + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "Google Nano Banana 2，最新文生图与编辑模型" + } + }, + { + "id": "gemini-3.1-pro-preview", + "displayName": "Gemini 3.1 Pro", + "family": "gemini-3.1", + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gemini-3-flash-preview", + "displayName": "Gemini 3 Flash", + "family": "gemini-3", + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gemini-2.5-pro", + "displayName": "Gemini 2.5 Pro", + "family": "gemini-2.5", + "match": { + "patterns": [ + "gemini-2.5-pro*" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gemini-2.5-flash", + "displayName": "Gemini 2.5 Flash", + "family": "gemini-2.5", + "match": { + "patterns": [ + "gemini-2.5-flash*" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gemini-3.1-flash-lite-preview", + "displayName": "Gemini 3.1 Flash-Lite Preview", + "family": "gemini-3.1", + "match": { + "exact": [ + "gemini-3.1-flash-lite-preview" + ], + "patterns": [ + "gemini-3.1-flash-lite*" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "Google Gemini 3.1 Flash-Lite Preview，低成本高吞吐多模态模型" + } + }, + { + "id": "text-embedding-005", + "displayName": "Text Embedding 005", + "family": "text-embedding", + "match": { + "exact": [ + "text-embedding-005" + ] + }, + "spec": { + "contextWindow": 2048, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "serviceType": [ + "embedding" + ], + "description": "Google 文本嵌入模型，768 维度" + } + } + ] +} diff --git a/compute/model-specs/happyhorse.json b/compute/model-specs/happyhorse.json new file mode 100644 index 0000000..7df4ee2 --- /dev/null +++ b/compute/model-specs/happyhorse.json @@ -0,0 +1,184 @@ +{ + "description": "HappyHorse（快乐马）视频生成系列模型规格。通过阿里云百炼平台提供推理服务。官网：https://happyhorse.com。所有模型均为视频生成（video_gen），无上下文窗口概念。", + "specs": [ + { + "id": "happyhorse-1.1-t2v", + "displayName": "HappyHorse 1.1 T2V", + "family": "happyhorse-t2v", + "match": { + "exact": [ + "happyhorse-1.1-t2v" + ], + "patterns": [ + "happyhorse-1.1-t2v*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 文生视频，进一步提升文本词义理解、镜头调度与动态生成表现" + }, + "routing": { + "tier": "flagship", + "routingPriority": 116, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "happyhorse-1.0-t2v", + "displayName": "HappyHorse 1.0 T2V", + "family": "happyhorse-t2v", + "match": { + "exact": [ + "happyhorse-1.0-t2v" + ], + "patterns": [ + "happyhorse-1.0-t2v*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.0 文生视频，具备高度还原的动态画面生成能力" + } + }, + { + "id": "happyhorse-1.1-i2v", + "displayName": "HappyHorse 1.1 I2V", + "family": "happyhorse-i2v", + "match": { + "exact": [ + "happyhorse-1.1-i2v" + ], + "patterns": [ + "happyhorse-1.1-i2v*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 图生视频，进一步提升画面质感、动态表现与跨片段一致性" + }, + "routing": { + "tier": "flagship", + "routingPriority": 115, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "happyhorse-1.0-i2v", + "displayName": "HappyHorse 1.0 I2V", + "family": "happyhorse-i2v", + "match": { + "exact": [ + "happyhorse-1.0-i2v" + ], + "patterns": [ + "happyhorse-1.0-i2v*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.0 图生视频，具备高度还原的动态画面生成能力" + } + }, + { + "id": "happyhorse-1.1-r2v", + "displayName": "HappyHorse 1.1 R2V", + "family": "happyhorse-r2v", + "match": { + "exact": [ + "happyhorse-1.1-r2v" + ], + "patterns": [ + "happyhorse-1.1-r2v*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "reference_to_video", + "multi_image", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 参考生视频，支持最多9张图片参考，更加稳定的主体与场景参考" + }, + "routing": { + "tier": "flagship", + "routingPriority": 117, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "happyhorse-1.0-r2v", + "displayName": "HappyHorse 1.0 R2V", + "family": "happyhorse-r2v", + "match": { + "exact": [ + "happyhorse-1.0-r2v" + ], + "patterns": [ + "happyhorse-1.0-r2v*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "reference_to_video", + "multi_image" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.0 参考生视频，支持最多9张图片参考" + } + } + ] +} diff --git a/compute/model-specs/kling.json b/compute/model-specs/kling.json new file mode 100644 index 0000000..8ff6d1a --- /dev/null +++ b/compute/model-specs/kling.json @@ -0,0 +1,133 @@ +{ + "description": "快手可灵视频生成系列模型规格。参数来源：config-center compute/providers/kling.json。所有模型均为视频生成（video_gen），无上下文窗口概念。", + "specs": [ + { + "id": "kling-3.0-turbo", + "displayName": "可灵 3.0 Turbo", + "family": "kling-3.0-turbo", + "match": { + "exact": [ + "kling-3.0-turbo" + ] + }, + "spec": { + "serviceType": [ + "video_gen" + ], + "capabilities": [ + "video_generation", + "image_to_video" + ], + "description": "可灵 3.0 Turbo，新版 /image-to-video/kling-3.0-turbo API，需独立协议适配" + } + }, + { + "id": "kling-v3", + "displayName": "可灵 3.0", + "family": "kling-v3", + "match": { + "exact": [ + "kling-v3" + ] + }, + "spec": { + "serviceType": [ + "video_gen" + ], + "capabilities": [ + "video_generation", + "image_to_video" + ], + "description": "可灵 3.0，通过现有 /v1/videos/text2video 任务 API 的 model_name 调用" + } + }, + { + "id": "kling-v2-5-turbo", + "displayName": "可灵 V2.5 Turbo", + "family": "kling-v2-5", + "match": { + "exact": [ + "kling-v2-5-turbo" + ], + "patterns": [ + "kling-v2-5-turbo*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "high_quality", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵最新视频生成模型，高性价比" + } + }, + { + "id": "kling-v2-5-turbo-pro", + "displayName": "可灵 V2.5 Turbo Pro", + "family": "kling-v2-5", + "match": { + "exact": [ + "kling-v2-5-turbo-pro" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "ultra_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵高品质视频生成，支持 4K 分辨率" + } + }, + { + "id": "kling-v2", + "displayName": "可灵 V2", + "family": "kling-v2", + "match": { + "exact": [ + "kling-v2" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵文生视频旗舰模型" + } + }, + { + "id": "kling-v2-master", + "displayName": "可灵 V2 Master", + "family": "kling-v2", + "match": { + "exact": [ + "kling-v2-master" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "ultra_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵最高品质视频生成，支持 4K 分辨率" + } + } + ] +} diff --git a/compute/model-specs/minimax.json b/compute/model-specs/minimax.json new file mode 100644 index 0000000..d0a59df --- /dev/null +++ b/compute/model-specs/minimax.json @@ -0,0 +1,762 @@ +{ + "description": "MiniMax 系列模型规格。参数来源：config-center compute/providers/minimax.json。id 保留上游 model_name 原始写法（含大小写与点号），匹配时由匹配器归一化处理。", + "specs": [ + { + "id": "MiniMax-H3-Max", + "displayName": "MiniMax H3 Max", + "family": "MiniMax-H3-Max", + "match": { + "exact": [ + "MiniMax-H3-Max" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "MiniMax H3 Max，使用原生视频 V2 API（/v1/video_generation_v2），不适用 Anthropic Messages", + "extra": { + "maxVideoDuration": 15 + } + } + }, + { + "id": "MiniMax-H3", + "displayName": "MiniMax H3", + "family": "MiniMax-H3", + "match": { + "exact": [ + "MiniMax-H3" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "MiniMax H3，使用原生视频 V2 API（/v1/video_generation_v2），不适用 Anthropic Messages", + "extra": { + "maxVideoDuration": 15 + } + } + }, + { + "id": "MiniMax-M3.1-Flash-Preview", + "displayName": "MiniMax M3.1 Flash Preview", + "family": "MiniMax-M3.1-Flash-Preview", + "match": { + "exact": [ + "MiniMax-M3.1-Flash-Preview" + ] + }, + "spec": { + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "MiniMax M3.1 Flash Preview", + "contextWindow": 1000000, + "extra": { + "adaptiveThinking": true, + "thinkingOnly": true + }, + "supportsReasoning": true + } + }, + { + "id": "MiniMax-M3", + "displayName": "MiniMax M3", + "family": "minimax-m3", + "match": { + "exact": [ + "MiniMax-M3" + ], + "patterns": [ + "MiniMax-M3*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 512000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "MiniMax M3 全模态基础模型，100 万上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "MiniMax-M2.7", + "displayName": "MiniMax M2.7", + "family": "minimax-m2.7", + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "MiniMax-M2.5", + "displayName": "MiniMax M2.5", + "family": "minimax-m2.5", + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "MiniMax-M2", + "displayName": "MiniMax M2", + "family": "minimax-m2", + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1 + } + }, + { + "id": "MiniMax-M2.7-highspeed", + "displayName": "MiniMax M2.7 高速版", + "family": "minimax-m2.7", + "match": { + "exact": [ + "MiniMax-M2.7-highspeed" + ] + }, + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "MiniMax M2.7 极速版，效果一致，输出更快，适合低延迟场景" + } + }, + { + "id": "MiniMax-M2.5-highspeed", + "displayName": "MiniMax M2.5 高速版", + "family": "minimax-m2.5", + "match": { + "exact": [ + "MiniMax-M2.5-highspeed" + ] + }, + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "MiniMax M2.5 高速推理版本（100TPS），适合低延迟场景" + } + }, + { + "id": "MiniMax-M2.1", + "displayName": "MiniMax M2.1", + "family": "minimax-m2.1", + "match": { + "exact": [ + "MiniMax-M2.1" + ] + }, + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "code", + "tool_use", + "reasoning" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "MiniMax 上一代主力模型，200K 上下文，性价比优秀" + } + }, + { + "id": "MiniMax-M2.1-highspeed", + "displayName": "MiniMax M2.1 高速版", + "family": "minimax-m2.1", + "match": { + "exact": [ + "MiniMax-M2.1-highspeed" + ] + }, + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "MiniMax M2.1 高速推理版本，适合低延迟场景" + } + }, + { + "id": "M2-her", + "displayName": "MiniMax M2 Her", + "family": "minimax-m2", + "match": { + "exact": [ + "M2-her" + ] + }, + "spec": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "roleplay" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "MiniMax 角色扮演对话模型，专为多轮角色对话设计" + } + }, + { + "id": "MiniMax-Text-01", + "displayName": "MiniMax Text 01", + "family": "minimax-text", + "match": { + "exact": [ + "MiniMax-Text-01" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.1, + "description": "MiniMax 百万级长上下文文本模型，擅长长文档处理" + } + }, + { + "id": "image-01", + "displayName": "MiniMax Image 01", + "family": "minimax-image", + "match": { + "exact": [ + "image-01" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "subject_reference", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图/图生图模型，支持主题参考（角色一致性）" + }, + "routing": { + "tier": "balanced", + "routingPriority": 113, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "image-01-live", + "displayName": "MiniMax Image 01 Live", + "family": "minimax-image", + "match": { + "exact": [ + "image-01-live" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "style_control", + "subject_reference", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图模型（Live 版），支持画风控制" + } + }, + { + "id": "speech-2.8-hd", + "displayName": "MiniMax Speech 2.8 HD", + "family": "minimax-speech", + "match": { + "exact": [ + "speech-2.8-hd" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 最新高清语音合成，40+ 语言、100+ 预置音色" + } + }, + { + "id": "speech-2.8-turbo", + "displayName": "MiniMax Speech 2.8 Turbo", + "family": "minimax-speech", + "match": { + "exact": [ + "speech-2.8-turbo" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成模型，速度优先，适合实时场景" + } + }, + { + "id": "speech-2.6-hd", + "displayName": "MiniMax Speech 2.6 HD", + "family": "minimax-speech", + "match": { + "exact": [ + "speech-2.6-hd" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 高清语音合成（2.6 版），音质优秀，生成速度快" + } + }, + { + "id": "speech-2.6-turbo", + "displayName": "MiniMax Speech 2.6 Turbo", + "family": "minimax-speech", + "match": { + "exact": [ + "speech-2.6-turbo" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成（2.6 版），低延迟响应" + } + }, + { + "id": "speech-02-hd", + "displayName": "MiniMax Speech 02 HD", + "family": "minimax-speech", + "match": { + "exact": [ + "speech-02-hd" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "voice_clone", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 高清语音合成（历史版本），韵律稳定性优秀" + } + }, + { + "id": "speech-02-turbo", + "displayName": "MiniMax Speech 02 Turbo", + "family": "minimax-speech", + "match": { + "exact": [ + "speech-02-turbo" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成（历史版本），增强小语种支持" + } + }, + { + "id": "MiniMax-Hailuo-2.3", + "displayName": "海螺视频 2.3", + "family": "minimax-hailuo", + "match": { + "exact": [ + "MiniMax-Hailuo-2.3" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "camera_control", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "MiniMax 最新视频生成模型，支持 15 种运镜指令控制" + }, + "routing": { + "tier": "balanced", + "routingPriority": 102, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "MiniMax-Hailuo-2.3-Fast", + "displayName": "海螺视频 2.3 快速版", + "family": "minimax-hailuo-2.3-fast", + "match": { + "exact": [ + "MiniMax-Hailuo-2.3-Fast", + "MiniMax-Hailuo-2.3-fast" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "camera_control", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "海螺视频快速版本（仅图生视频），生成速度更快" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 101, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "T2V-01-Director", + "displayName": "海螺视频 Director", + "family": "minimax-hailuo", + "match": { + "exact": [ + "T2V-01-Director" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video", + "camera_control", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "导演级文生视频模型，精准运镜控制（历史版本）" + } + }, + { + "id": "MiniMax-Hailuo-02", + "displayName": "海螺视频 02", + "family": "minimax-hailuo", + "match": { + "exact": [ + "MiniMax-Hailuo-02" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "海螺视频 02，原生 1080P 生成（历史版本）" + } + }, + { + "id": "S2V-01", + "displayName": "海螺视频 主体参考", + "family": "minimax-hailuo", + "match": { + "exact": [ + "S2V-01" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "subject_reference", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "主体参考视频生成，基于参考图保持角色一致性" + } + }, + { + "id": "music-2.6", + "displayName": "MiniMax Music 2.6", + "family": "minimax-music", + "match": { + "exact": [ + "music-2.6" + ] + }, + "spec": { + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成模型，支持音乐灵感描述和歌词输入；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + } + }, + { + "id": "music-2.5+", + "displayName": "MiniMax Music 2.5+", + "family": "minimax-music", + "match": { + "exact": [ + "music-2.5+" + ] + }, + "spec": { + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成（2.5+ 版），音质提升；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + } + }, + { + "id": "music-2.5", + "displayName": "MiniMax Music 2.5", + "family": "minimax-music", + "match": { + "exact": [ + "music-2.5" + ] + }, + "spec": { + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成（2.5 版）；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + } + }, + { + "id": "music-cover", + "displayName": "MiniMax Music Cover", + "family": "minimax-music", + "match": { + "exact": [ + "music-cover" + ] + }, + "spec": { + "capabilities": [ + "music_generation", + "cover", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 翻唱模型，基于参考音频生成翻唱版本；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + } + } + ] +} diff --git a/compute/model-specs/mistral.json b/compute/model-specs/mistral.json new file mode 100644 index 0000000..244e2de --- /dev/null +++ b/compute/model-specs/mistral.json @@ -0,0 +1,107 @@ +{ + "description": "Mistral 系列模型规格。参数来源：config-center compute/providers/mistral.json。", + "specs": [ + { + "id": "mistral-medium-3.5", + "displayName": "Mistral Medium 3.5", + "family": "mistral-medium-3-5", + "match": { + "exact": [ + "mistral-medium-3-5", + "mistral-medium-3.5", + "mistral-medium-latest" + ] + }, + "spec": { + "contextWindow": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Mistral Medium 3.5，官方 API ID mistral-medium-3-5，256K 上下文" + } + }, + { + "id": "mistral-large-latest", + "displayName": "Mistral Large", + "family": "mistral-large-2512", + "match": { + "exact": [ + "mistral-large-latest", + "mistral-large-2512" + ] + }, + "spec": { + "contextWindow": 256000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + } + }, + { + "id": "mistral-small-latest", + "displayName": "Mistral Small 4", + "family": "mistral-small-2603", + "match": { + "exact": [ + "mistral-small-latest", + "mistral-small-2603" + ] + }, + "spec": { + "contextWindow": 262144, + "capabilities": [ + "chat", + "code", + "vision", + "reasoning", + "fast", + "tool_use" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 0.7 + } + }, + { + "id": "codestral-latest", + "displayName": "Codestral", + "family": "codestral", + "match": { + "patterns": [ + "codestral*" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 32768, + "capabilities": [ + "code", + "fill_in_the_middle" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + } + } + ] +} diff --git a/compute/model-specs/moonshot.json b/compute/model-specs/moonshot.json new file mode 100644 index 0000000..79c1911 --- /dev/null +++ b/compute/model-specs/moonshot.json @@ -0,0 +1,255 @@ +{ + "description": "Moonshot Kimi 系列模型规格。参数来源：config-center compute/providers/moonshot.json。注意 kimi-k2 存在多个子版本（k2.5/k2.6/k2-thinking），故 kimi-k2 仅用 exact 主键匹配，不用宽 pattern。", + "specs": [ + { + "id": "kimi-k3", + "displayName": "Kimi K3", + "family": "kimi-k3", + "match": { + "exact": [ + "kimi-k3", + "moonshotai/kimi-k3" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 1048576, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding" + ], + "serviceType": [ + "reasoning" + ], + "supportsReasoning": true, + "description": "Kimi K3，1M 上下文，始终开启思考，low/high/max 三档（默认 max）；输出上限与输入共享窗口，默认 max_completion_tokens=131072", + "extra": { + "thinkingOnly": true, + "samplingParametersDeprecated": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + } + }, + { + "id": "kimi-k2.7-code", + "displayName": "Kimi K2.7 Code", + "family": "kimi-k2.7", + "match": { + "exact": [ + "kimi-k2.7-code" + ], + "patterns": [ + "kimi-k2.7-code*" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Moonshot Kimi K2.7 Code 编程专用模型" + }, + "routing": { + "tier": "balanced", + "routingPriority": 50, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "kimi-k2.6", + "displayName": "Kimi K2.6", + "family": "kimi-k2.6", + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 45, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "kimi-k2.5", + "displayName": "Kimi K2.5", + "family": "kimi-k2.5", + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "kimi-k2-thinking", + "displayName": "Kimi K2 Thinking", + "family": "kimi-k2-thinking", + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 16384, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "kimi-k2", + "displayName": "Kimi K2", + "family": "kimi-k2", + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true + } + }, + { + "id": "moonshot-v1-8k", + "displayName": "Moonshot v1 8K", + "family": "moonshot-v1", + "spec": { + "contextWindow": 8192, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0 + } + }, + { + "id": "moonshot-v1-32k", + "displayName": "Moonshot v1 32K", + "family": "moonshot-v1", + "spec": { + "contextWindow": 32768, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0 + } + }, + { + "id": "moonshot-v1-128k", + "displayName": "Moonshot v1 128K", + "family": "moonshot-v1", + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0 + } + } + ] +} diff --git a/compute/model-specs/openai.json b/compute/model-specs/openai.json new file mode 100644 index 0000000..b40b4e1 --- /dev/null +++ b/compute/model-specs/openai.json @@ -0,0 +1,1324 @@ +{ + "description": "OpenAI GPT / o 系列模型规格。参数来源：config-center compute/providers/openai.json。注意：gpt-4o / gpt-5 等存在子型号（-mini/-nano），故只用 exact 主键匹配，不使用宽泛 pattern 以防误吞子型号。o 系列 serviceType 由 responses 调整为 reasoning（responses 为 OpenAI 直连专有，对 NewAPI 网关不适用）。", + "specs": [ + { + "id": "gpt-6.1-sol", + "displayName": "GPT-6.1 Sol", + "family": "gpt-6.1-sol", + "match": { + "exact": [ + "gpt-6.1-sol", + "openai/gpt-6.1-sol" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-6.1 Sol，1.05M 上下文，128K 最大输出；工具调用需要 Responses API，不支持关闭推理" + }, + "routing": { + "tier": "balanced", + "routingPriority": 4, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-image-2.5-sunburst", + "displayName": "GPT Image 2.5 Sunburst", + "family": "gpt-image-2.5-sunburst", + "match": { + "exact": [ + "gpt-image-2.5-sunburst" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "图像生成与编辑模型，适合对编辑精度要求高的任务" + } + }, + { + "id": "gpt-image-2.5-flare", + "displayName": "GPT Image 2.5 Flare", + "family": "gpt-image-2.5-flare", + "match": { + "exact": [ + "gpt-image-2.5-flare" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision", + "fast" + ], + "serviceType": [ + "image_gen" + ], + "description": "快速图像生成与编辑模型，适合日常高质量出图" + } + }, + { + "id": "gpt-6-sol", + "displayName": "GPT-6 Sol", + "family": "gpt-6-sol", + "match": { + "exact": [ + "gpt-6-sol" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-6 Sol，1.05M 上下文，128K 最大输出" + }, + "routing": { + "tier": "balanced", + "routingPriority": 5, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-6-luna", + "displayName": "GPT-6 Luna", + "family": "gpt-6-luna", + "match": { + "exact": [ + "gpt-6-luna" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-6 Luna，1.05M 上下文，128K 最大输出" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 5, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-6-astra", + "displayName": "GPT-6 Astra", + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "extra": { + "thinkingOnly": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-reserve", + "displayName": "GPT-Reserve", + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "参数暂按 GPT-5.6 Luna 配置，待官方规格确认；保留手动选择，未经官方规格核验前不参与 Agent 自动选型" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 10, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-image-2", + "displayName": "GPT Image 2", + "family": "gpt-image", + "match": { + "exact": [ + "gpt-image-2" + ] + }, + "spec": { + "contextWindow": 400000, + "capabilities": [ + "image_generation", + "image_editing", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "OpenAI GPT Image 2 文生图模型，支持高质量图片生成与编辑，400K 上下文" + } + }, + { + "id": "gpt-5.5", + "displayName": "GPT-5.5", + "family": "gpt-5", + "match": { + "exact": [ + "gpt-5.5" + ], + "patterns": [ + "gpt-5.5*" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "flagship", + "routingPriority": 60, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-5.4", + "displayName": "GPT-5.4", + "family": "gpt-5", + "match": { + "exact": [ + "gpt-5.4", + "gpt-5.4-2026-03-05" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 60, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-5.2", + "displayName": "GPT-5.2", + "family": "gpt-5", + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gpt-5.1", + "displayName": "GPT-5.1", + "family": "gpt-5", + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gpt-5", + "displayName": "GPT-5", + "family": "gpt-5", + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gpt-5-mini", + "displayName": "GPT-5 Mini", + "family": "gpt-5", + "match": { + "patterns": [ + "gpt-5-mini*" + ] + }, + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "fast", + "tool_use", + "code", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "gpt-5-nano", + "displayName": "GPT-5 Nano", + "family": "gpt-5", + "match": { + "patterns": [ + "gpt-5-nano*" + ] + }, + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "code", + "fast", + "tool_use" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1 + } + }, + { + "id": "gpt-4.1", + "displayName": "GPT-4.1", + "family": "gpt-4.1", + "spec": { + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1 + } + }, + { + "id": "gpt-4o", + "displayName": "GPT-4o", + "family": "gpt-4o", + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1 + } + }, + { + "id": "gpt-4o-mini", + "displayName": "GPT-4o Mini", + "family": "gpt-4o", + "match": { + "patterns": [ + "gpt-4o-mini*" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1 + } + }, + { + "id": "o3", + "displayName": "o3", + "family": "o3", + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + } + }, + { + "id": "o3-mini", + "displayName": "o3 Mini", + "family": "o3", + "match": { + "patterns": [ + "o3-mini*" + ] + }, + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "math", + "code", + "fast" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + } + }, + { + "id": "o4-mini", + "displayName": "o4 Mini", + "family": "o4", + "match": { + "patterns": [ + "o4-mini*" + ] + }, + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "math", + "code", + "tool_use", + "vision" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + } + }, + { + "id": "gpt-5.2-pro", + "displayName": "GPT-5.2 Pro", + "family": "gpt-5", + "match": { + "exact": [ + "gpt-5.2-pro" + ] + }, + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "OpenAI GPT-5.2 Pro 专业版，最高性能" + } + }, + { + "id": "gpt-5-pro", + "displayName": "GPT-5 Pro", + "family": "gpt-5", + "match": { + "exact": [ + "gpt-5-pro" + ] + }, + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 272000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "OpenAI GPT-5 Pro 专业版高性能推理模型" + } + }, + { + "id": "gpt-4.1-mini", + "displayName": "GPT-4.1 mini", + "family": "gpt-4.1", + "match": { + "exact": [ + "gpt-4.1-mini" + ], + "patterns": [ + "gpt-4.1-mini*" + ] + }, + "spec": { + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "OpenAI GPT-4.1 mini，1M 上下文，高性价比" + } + }, + { + "id": "gpt-4.1-nano", + "displayName": "GPT-4.1 nano", + "family": "gpt-4.1", + "match": { + "exact": [ + "gpt-4.1-nano" + ], + "patterns": [ + "gpt-4.1-nano*" + ] + }, + "spec": { + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "code", + "ultra_long_context", + "fast", + "tool_use" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "OpenAI GPT-4.1 nano，1M 上下文，极致性价比" + } + }, + { + "id": "text-embedding-3-small", + "displayName": "text-embedding-3-small", + "family": "text-embedding-3", + "match": { + "exact": [ + "text-embedding-3-small" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "serviceType": [ + "embedding" + ], + "description": "OpenAI 轻量文本向量化模型，1536 维" + } + }, + { + "id": "text-embedding-3-large", + "displayName": "text-embedding-3-large", + "family": "text-embedding-3", + "match": { + "exact": [ + "text-embedding-3-large" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "serviceType": [ + "embedding" + ], + "description": "OpenAI 高精度文本向量化模型，3072 维" + } + }, + { + "id": "tts-1", + "displayName": "TTS-1", + "family": "tts-1", + "match": { + "exact": [ + "tts-1" + ] + }, + "spec": { + "capabilities": [ + "tts" + ], + "serviceType": [ + "tts" + ], + "description": "OpenAI 标准语音合成" + } + }, + { + "id": "tts-1-hd", + "displayName": "TTS-1-HD", + "family": "tts-1", + "match": { + "exact": [ + "tts-1-hd" + ] + }, + "spec": { + "capabilities": [ + "tts" + ], + "serviceType": [ + "tts" + ], + "description": "OpenAI 高清语音合成" + } + }, + { + "id": "whisper-1", + "displayName": "Whisper", + "family": "whisper", + "match": { + "exact": [ + "whisper-1" + ] + }, + "spec": { + "capabilities": [ + "asr", + "multilingual" + ], + "serviceType": [ + "asr" + ], + "description": "OpenAI 通用语音识别" + } + }, + { + "id": "o3-pro", + "displayName": "o3-pro", + "family": "o3", + "match": { + "exact": [ + "o3-pro" + ] + }, + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "math", + "science" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "OpenAI o3-pro 高级推理模型，200K 上下文" + } + }, + { + "id": "dall-e-3", + "displayName": "DALL-E 3", + "family": "dall-e", + "match": { + "exact": [ + "dall-e-3" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "OpenAI 高质量文生图模型" + } + }, + { + "id": "gpt-4o-realtime", + "displayName": "GPT-4o Realtime", + "family": "gpt-4o-realtime", + "match": { + "exact": [ + "gpt-4o-realtime" + ] + }, + "spec": { + "contextWindow": 32000, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "realtime" + ], + "serviceType": [ + "omni" + ], + "defaultTemperature": 0.7, + "description": "OpenAI 全能实时模型，支持语音 + 文本 + 视觉" + } + }, + { + "id": "gpt-4o-realtime-preview", + "displayName": "GPT-4o Realtime Preview", + "family": "gpt-4o-realtime", + "match": { + "exact": [ + "gpt-4o-realtime-preview" + ], + "patterns": [ + "gpt-4o-realtime-preview*" + ] + }, + "spec": { + "contextWindow": 32000, + "maxOutputTokens": 4096, + "capabilities": [ + "realtime", + "speech_to_speech", + "low_latency" + ], + "serviceType": [ + "realtime_voice" + ], + "defaultTemperature": 0.7, + "description": "OpenAI 实时语音交互模型（预览版）" + } + }, + { + "id": "gpt-oss-120b", + "displayName": "GPT OSS 120B", + "family": "gpt-oss", + "match": { + "exact": [ + "gpt-oss-120b", + "openai/gpt-oss-120b:free", + "gpt-oss-120b:free" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "reasoning", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "description": "OpenAI 开源 120B 模型（通过 OpenRouter 提供，:free 层）" + } + }, + { + "id": "gpt-5.6-sol", + "displayName": "GPT-5.6 Sol", + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "flagship", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "low" + } + } + }, + { + "id": "gpt-5.6-sol-pro", + "displayName": "GPT-5.6 Sol Pro", + "family": "gpt-5.6-sol-pro", + "match": { + "exact": [ + "gpt-5.6-sol-pro" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-5.6 Sol 的 Pro 固定推理变体，由上游以 reasoning.mode=pro 提供", + "releasedAt": "2026-07-09" + } + }, + { + "id": "gpt-5.6-terra", + "displayName": "GPT-5.6 Terra", + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-5.6-terra-pro", + "displayName": "GPT-5.6 Terra Pro", + "family": "gpt-5.6-terra-pro", + "match": { + "exact": [ + "gpt-5.6-terra-pro" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-5.6 Terra 的 Pro 固定推理变体，由上游以 reasoning.mode=pro 提供", + "releasedAt": "2026-07-09" + } + }, + { + "id": "gpt-5.6-luna", + "displayName": "GPT-5.6 Luna", + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "lightweight", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "gpt-5.6-luna-pro", + "displayName": "GPT-5.6 Luna Pro", + "family": "gpt-5.6-luna-pro", + "match": { + "exact": [ + "gpt-5.6-luna-pro" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-5.6 Luna 的 Pro 固定推理变体，由上游以 reasoning.mode=pro 提供", + "releasedAt": "2026-07-09" + } + }, + { + "id": "gpt-5.4-mini", + "displayName": "GPT-5.4 Mini", + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "lightweight", + "routingPriority": 25, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "low" + } + } + }, + { + "id": "gpt-5.3-codex-spark", + "displayName": "GPT-5.3 Codex Spark（Pro 专属）", + "spec": { + "contextWindow": 128000, + "capabilities": [ + "chat", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ] + }, + "routing": { + "tier": "lightweight", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "gpt-5.4-nano", + "displayName": "GPT-5.4 Nano", + "family": "gpt-5.4-nano", + "match": { + "exact": [ + "gpt-5.4-nano", + "gpt-5.4-nano-2026-03-17" + ] + }, + "spec": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "description": "GPT-5.4 Nano；官网确认400000总窗口、128000最大输出。2026-10-01宣布deprecated，停止服务日2027-04-01；当前不据此退役。独立输入上限未由本页确认。" + } + }, + { + "id": "gpt-realtime-2.1", + "displayName": "gpt-realtime-2.1", + "family": "gpt-realtime-2.1", + "match": { + "exact": [ + "gpt-realtime-2.1" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 32000, + "description": "官方精确型号与窗口/输出已核；独立实时协议适配与账号验收由接入面决定，不继承通用Chat合同。" + } + }, + { + "id": "gpt-realtime-2.1-mini", + "displayName": "gpt-realtime-2.1-mini", + "family": "gpt-realtime-2.1-mini", + "match": { + "exact": [ + "gpt-realtime-2.1-mini" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 32000, + "description": "官方精确型号与窗口/输出已核；独立实时协议适配与账号验收由接入面决定，不继承通用Chat合同。" + } + }, + { + "id": "gpt-realtime-translate", + "displayName": "gpt-realtime-translate", + "family": "gpt-realtime-translate", + "match": { + "exact": [ + "gpt-realtime-translate" + ] + }, + "spec": { + "contextWindow": 16000, + "maxOutputTokens": 2000, + "description": "官方精确型号与窗口/输出已核；独立实时协议适配与账号验收由接入面决定，不继承通用Chat合同。" + } + }, + { + "id": "gpt-5.4-pro", + "displayName": "GPT-5.4 Pro", + "family": "gpt-5.4-pro", + "match": { + "exact": [ + "gpt-5.4-pro" + ] + }, + "spec": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "description": "官方独立Pro规格；Responses协议限制由实际接入面执行，不沿用基础型号或其他代际family。" + } + } + ] +} diff --git a/compute/model-specs/perplexity.json b/compute/model-specs/perplexity.json new file mode 100644 index 0000000..3505837 --- /dev/null +++ b/compute/model-specs/perplexity.json @@ -0,0 +1,85 @@ +{ + "description": "Perplexity Sonar 系列模型规格。参数来源：config-center compute/providers/perplexity.json。所有模型均内置实时联网搜索能力。", + "specs": [ + { + "id": "sonar-pro", + "displayName": "Sonar Pro", + "family": "sonar-pro", + "match": { + "exact": [ + "sonar-pro" + ] + }, + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "web_search", + "reasoning", + "citation" + ], + "serviceType": [ + "chat" + ], + "description": "Perplexity 旗舰搜索增强模型，内置实时联网搜索和引用，200K 上下文" + } + }, + { + "id": "sonar-reasoning-pro", + "displayName": "Sonar Reasoning Pro", + "family": "sonar-reasoning", + "match": { + "exact": [ + "sonar-reasoning-pro" + ], + "patterns": [ + "sonar-reasoning*" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "web_search", + "reasoning", + "deep_thinking" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "Perplexity 深度推理模型，内置联网搜索和深度思考" + } + }, + { + "id": "sonar", + "displayName": "Sonar", + "family": "sonar", + "match": { + "exact": [ + "sonar" + ], + "patterns": [ + "sonar-*" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "web_search", + "citation", + "fast" + ], + "serviceType": [ + "chat" + ], + "description": "Perplexity 轻量搜索模型，低成本联网搜索" + } + } + ] +} diff --git a/compute/model-specs/qwen.json b/compute/model-specs/qwen.json new file mode 100644 index 0000000..e762947 --- /dev/null +++ b/compute/model-specs/qwen.json @@ -0,0 +1,1188 @@ +{ + "description": "阿里通义千问 Qwen 系列模型规格。参数来源：config-center compute/providers/dashscope.json。", + "specs": [ + { + "id": "qwen3.7-max", + "displayName": "Qwen3.7 Max", + "family": "qwen3.7-max", + "match": { + "exact": [ + "qwen3.7-max", + "qwen3.7-max-2026-05-20" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Qwen3.7 Max，纯文本模型，100 万上下文，131072 最大输出；当前版本等同 2026-05-20 快照" + } + }, + { + "id": "qwen3.8-max-prime", + "displayName": "Qwen3.8 Max Prime", + "family": "qwen3.8-max-prime", + "match": { + "exact": [ + "qwen3.8-max-prime", + "qwen/qwen3.8-max-prime" + ] + }, + "spec": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Qwen3.8 Max Prime；OpenRouter 官方目录已提供该 ID，不推断国内 DashScope 同名接入" + } + }, + { + "id": "qwen-image-3.0", + "displayName": "qwen-image-3.0", + "family": "qwen-image-3.0", + "match": { + "exact": [ + "qwen-image-3.0" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0" + } + }, + { + "id": "qwen-image-3.0-pro", + "displayName": "qwen-image-3.0-pro", + "family": "qwen-image-3.0-pro", + "match": { + "exact": [ + "qwen-image-3.0-pro" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0-pro" + } + }, + { + "id": "qwen3.7-text-rerank", + "displayName": "Qwen3.7 Text Rerank", + "family": "qwen3.7-text-rerank", + "match": { + "exact": [ + "qwen3.7-text-rerank" + ] + }, + "spec": { + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Qwen3.7 Text Rerank" + } + }, + { + "id": "qwen3.7-text-embedding-flash", + "displayName": "qwen3.7-text-embedding-flash", + "family": "qwen3.7-text-embedding-flash", + "match": { + "exact": [ + "qwen3.7-text-embedding-flash" + ] + }, + "spec": { + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "qwen3.7-text-embedding-flash" + } + }, + { + "id": "qwen3.7-text-embedding", + "displayName": "qwen3.7-text-embedding", + "family": "qwen3.7-text-embedding", + "match": { + "exact": [ + "qwen3.7-text-embedding" + ] + }, + "spec": { + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "qwen3.7-text-embedding" + } + }, + { + "id": "qwen3.8-omni-flash", + "displayName": "Qwen3.8 Omni Flash", + "family": "qwen3.8-omni-flash", + "match": { + "exact": [ + "qwen3.8-omni-flash", + "qwen/qwen3.8-omni-flash" + ] + }, + "spec": { + "capabilities": [ + "chat", + "reasoning", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "Qwen3.8 Omni Flash", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "supportsReasoning": true + } + }, + { + "id": "qwen3.8-max", + "displayName": "Qwen3.8 Max", + "family": "qwen3.8-max", + "match": { + "exact": [ + "qwen3.8-max" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context", + "agent", + "vision", + "video_understanding" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "通义千问 Qwen3.8 Max 稳定版，100 万上下文，支持视觉与视频理解" + }, + "routing": { + "tier": "flagship", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "xhigh" + } + } + }, + { + "id": "qwen3.8-flash", + "displayName": "Qwen3.8 Flash", + "family": "qwen3.8-flash", + "match": { + "exact": [ + "qwen3.8-flash" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "multilingual", + "tool_use", + "long_context", + "agent", + "vision", + "video_understanding", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "supportsReasoning": true, + "extra": { + "thinkingDefault": true, + "thinkingMaxTokens": 262144 + }, + "releasedAt": "2026-08-26", + "description": "通义千问 Qwen3.8 Flash，高性价比多模态推理模型，支持 100 万上下文及文本、图像、视频输入" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "xhigh" + } + } + }, + { + "id": "qwen3.8-max-preview", + "displayName": "Qwen3.8 Max Preview", + "family": "qwen3.8-max-preview", + "match": { + "exact": [ + "qwen3.8-max-preview" + ], + "patterns": [ + "qwen3.8-max-preview*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context", + "agent", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "通义千问 Qwen3.8 Max 预览版，100 万上下文，支持视觉理解" + }, + "routing": { + "tier": "flagship", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + } + }, + { + "id": "qwen3.7-plus", + "displayName": "Qwen3.7 Plus", + "family": "qwen3.7", + "match": { + "exact": [ + "qwen3.7-plus" + ], + "patterns": [ + "qwen3.7-plus*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "通义千问 Qwen3.7 Plus 高性价比多模态模型，100 万上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "qwen-max", + "displayName": "Qwen Max", + "family": "qwen-max", + "match": { + "patterns": [ + "qwen-max*" + ] + }, + "spec": { + "contextWindow": 32768, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + } + }, + { + "id": "qwen-plus", + "displayName": "Qwen Plus", + "family": "qwen-plus", + "match": { + "patterns": [ + "qwen-plus*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "thinking_mode", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + } + }, + { + "id": "qwen-turbo", + "displayName": "Qwen Turbo", + "family": "qwen-turbo", + "match": { + "patterns": [ + "qwen-turbo*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "long_context", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + } + }, + { + "id": "qwen3-max", + "displayName": "Qwen3 Max", + "family": "qwen3-max", + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "supportsReasoning": true + } + }, + { + "id": "qwen3.5-plus", + "displayName": "Qwen3.5 Plus", + "family": "qwen3.5", + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true + } + }, + { + "id": "qwen3.6-plus", + "displayName": "Qwen3.6 Plus", + "family": "qwen3.6", + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true + } + }, + { + "id": "qwen3.6-flash", + "displayName": "Qwen3.6 Flash", + "family": "qwen3.6", + "match": { + "exact": [ + "qwen3.6-flash" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "fast" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true, + "description": "通义千问3.6 Flash，高性价比多模态模型，100万上下文" + } + }, + { + "id": "qwen-long", + "displayName": "Qwen Long", + "family": "qwen-long", + "match": { + "exact": [ + "qwen-long" + ] + }, + "spec": { + "contextWindow": 10000000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "ultra_long_context", + "document_processing" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "通义千问 Long，支持 1000 万上下文，适合超长文档处理" + } + }, + { + "id": "qwen3-vl-plus", + "displayName": "Qwen3-VL-Plus", + "family": "qwen3-vl", + "match": { + "exact": [ + "qwen3-vl-plus" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "ocr", + "chart_analysis", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 0.7, + "description": "通义千问3代视觉模型Plus版，支持图像理解" + } + }, + { + "id": "qwen3-vl-flash", + "displayName": "Qwen3-VL-Flash", + "family": "qwen3-vl", + "match": { + "exact": [ + "qwen3-vl-flash" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "fast", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 0.7, + "description": "通义千问3代视觉模型Flash版，高性价比" + } + }, + { + "id": "text-embedding-v3", + "displayName": "text-embedding-v3", + "family": "text-embedding-v3", + "match": { + "exact": [ + "text-embedding-v3" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "chinese_optimized" + ], + "serviceType": [ + "embedding" + ], + "description": "阿里云文本向量化模型 v3" + } + }, + { + "id": "text-embedding-v4", + "displayName": "text-embedding-v4", + "family": "text-embedding-v4", + "match": { + "exact": [ + "text-embedding-v4" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag" + ], + "serviceType": [ + "embedding" + ], + "description": "阿里云百炼文本向量模型v4，支持多维度与多语种" + } + }, + { + "id": "qwen3-rerank", + "displayName": "Qwen3-Rerank", + "family": "qwen3-rerank", + "match": { + "exact": [ + "qwen3-rerank" + ] + }, + "spec": { + "contextWindow": 120000, + "capabilities": [ + "rerank", + "semantic_reranking", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "检索重排序，支持 100+ 语种" + } + }, + { + "id": "cosyvoice-v2", + "displayName": "CosyVoice V2", + "family": "cosyvoice", + "match": { + "exact": [ + "cosyvoice-v2" + ] + }, + "spec": { + "capabilities": [ + "tts", + "multilingual", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "通义语音合成，支持中英文" + } + }, + { + "id": "paraformer-v2", + "displayName": "Paraformer V2", + "family": "paraformer", + "match": { + "exact": [ + "paraformer-v2" + ] + }, + "spec": { + "capabilities": [ + "asr", + "chinese_optimized" + ], + "serviceType": [ + "asr" + ], + "description": "通义语音识别" + } + }, + { + "id": "wan2.7-image-pro", + "displayName": "通义万相 2.7 Pro", + "family": "wan2", + "match": { + "exact": [ + "wan2.7-image-pro" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相最新旗舰文生图模型，支持 4K 分辨率" + }, + "routing": { + "tier": "flagship", + "routingPriority": 100, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "wan2.7-image", + "displayName": "通义万相 2.7", + "family": "wan2", + "match": { + "exact": [ + "wan2.7-image" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相最新标准文生图模型，高画质" + }, + "routing": { + "tier": "balanced", + "routingPriority": 107, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "wan2.6-t2i", + "displayName": "通义万相 2.6 文生图", + "family": "wan2", + "match": { + "exact": [ + "wan2.6-t2i" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相 2.6 文生图专用模型，稳定可靠" + } + }, + { + "id": "wan2.2-t2i-plus", + "displayName": "通义万相 2.2 Plus", + "family": "wan2", + "match": { + "exact": [ + "wan2.2-t2i-plus" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相 2.2 增强版，画质与速度平衡" + } + }, + { + "id": "wan2.2-t2i-flash", + "displayName": "通义万相 2.2 Flash", + "family": "wan2", + "match": { + "exact": [ + "wan2.2-t2i-flash" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized", + "fast" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相 2.2 快速版，生成速度极快" + } + }, + { + "id": "wan2.6-t2v", + "displayName": "通义万相 2.6 文生视频", + "family": "wan2", + "match": { + "exact": [ + "wan2.6-t2v" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "万相 2.6 文生视频模型，支持文生视频和图生视频" + } + }, + { + "id": "cosyvoice-clone", + "displayName": "CosyVoice 声音复刻", + "family": "cosyvoice", + "match": { + "exact": [ + "cosyvoice-clone" + ] + }, + "spec": { + "capabilities": [ + "voice_clone", + "chinese_optimized", + "few_shot" + ], + "serviceType": [ + "voice_clone" + ], + "description": "通义声音复刻，少量样本即可克隆" + } + }, + { + "id": "qwen-omni-turbo", + "displayName": "Qwen-Omni-Turbo", + "family": "qwen-omni", + "match": { + "exact": [ + "qwen-omni-turbo" + ] + }, + "spec": { + "contextWindow": 32768, + "maxOutputTokens": 2048, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "multimodal", + "tool_use" + ], + "serviceType": [ + "omni" + ], + "defaultTemperature": 0.7, + "description": "通义全能模型，支持音视频文本多模态" + } + }, + { + "id": "qwen3-max-trans", + "displayName": "Qwen3-Max（翻译）", + "family": "qwen3-trans", + "match": { + "exact": [ + "qwen3-max-trans" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 8192, + "capabilities": [ + "translation", + "multilingual", + "chinese_optimized" + ], + "serviceType": [ + "translation" + ], + "defaultTemperature": 0.65, + "description": "通义千问翻译，中英日韩等多语种互译" + } + }, + { + "id": "qwen3-coder", + "displayName": "Qwen3 Coder", + "family": "qwen3-coder", + "match": { + "exact": [ + "qwen3-coder", + "qwen/qwen3-coder:free", + "qwen3-coder:free" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "code", + "long_context", + "agent", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "通义千问3 Coder 编程模型（通过 OpenRouter 提供，:free 层）" + } + }, + { + "id": "qwen3.5-35b-a3b", + "displayName": "Qwen3.5 35B-A3B", + "family": "qwen3.5", + "match": { + "exact": [ + "qwen3.5-35b-a3b" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3.5 35B-A3B MoE 模型（35B 总参/3B 激活），131K 上下文" + } + }, + { + "id": "qwen3.5-27b", + "displayName": "Qwen3.5 27B", + "family": "qwen3.5", + "match": { + "exact": [ + "qwen3.5-27b" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3.5 27B Dense 模型，131K 上下文" + } + }, + { + "id": "qwen3-coder-480b", + "displayName": "Qwen3 Coder 480B (硅基)", + "family": "qwen3-coder", + "match": { + "exact": [ + "qwen3-coder-480b", + "Qwen/Qwen3-Coder-480B-A35B-Instruct" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "code", + "long_context", + "agent", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3-Coder-480B-A35B，480B MoE（35B 激活），256K 上下文，面向 Agentic Coding" + } + }, + { + "id": "qwen3-235b", + "displayName": "Qwen3 235B (硅基)", + "family": "qwen3-235b", + "match": { + "exact": [ + "qwen3-235b", + "Qwen/Qwen3-235B-A22B-Instruct-2507" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "code", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3-235B-A22B-Instruct-2507，235B MoE 通用指令模型，262K 上下文" + } + }, + { + "id": "qwen-image-2.0-pro", + "displayName": "千问图像 2.0 Pro", + "family": "qwen-image-2", + "match": { + "exact": [ + "qwen-image-2.0-pro" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "千问图像生成 2.0 Pro 旗舰版，高画质多风格" + } + }, + { + "id": "qwen-image-2.0", + "displayName": "千问图像 2.0", + "family": "qwen-image-2", + "match": { + "exact": [ + "qwen-image-2.0" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "千问图像生成 2.0 标准版" + } + }, + { + "id": "qwen3-coder-plus", + "displayName": "qwen3-coder-plus", + "family": "qwen3-coder-plus", + "match": { + "exact": [ + "qwen3-coder-plus" + ] + }, + "spec": { + "description": "精确绑定已有接入型号，参数未核实；identity-only规格不继承旧family能力、窗口、协议或自动路由资格。" + } + }, + { + "id": "qwen3-coder-next", + "displayName": "qwen3-coder-next", + "family": "qwen3-coder-next", + "match": { + "exact": [ + "qwen3-coder-next" + ] + }, + "spec": { + "description": "精确绑定已有接入型号，参数未核实；identity-only规格不继承旧family能力、窗口、协议或自动路由资格。" + } + } + ] +} diff --git a/compute/model-specs/stability.json b/compute/model-specs/stability.json new file mode 100644 index 0000000..cc209ee --- /dev/null +++ b/compute/model-specs/stability.json @@ -0,0 +1,50 @@ +{ + "description": "Stability AI 图像生成系列模型规格。参数来源：config-center compute/providers/stability.json。无上下文窗口（图像生成模型）。", + "specs": [ + { + "id": "stable-audio-3.0", + "displayName": "Stable Audio 3.0", + "family": "stable-audio-3.0", + "match": { + "exact": [ + "stable-audio-3.0" + ] + }, + "spec": { + "capabilities": [ + "music_generation", + "audio_generation" + ], + "serviceType": [ + "music_gen" + ], + "releasedAt": "2026-05-20", + "description": "Stable Audio 3.0，通过原生 /v2beta/audio/stable-audio API 提供，不使用 Stability 图像协议" + } + }, + { + "id": "stable-diffusion-3.5-large", + "displayName": "Stable Diffusion 3.5 Large", + "family": "stable-diffusion-3", + "match": { + "exact": [ + "stable-diffusion-3.5-large" + ], + "patterns": [ + "stable-diffusion-3.5*" + ] + }, + "spec": { + "capabilities": [ + "image_generation", + "high_quality", + "controllable" + ], + "serviceType": [ + "image_gen" + ], + "description": "Stability AI 旗舰文生图模型，支持多种分辨率和风格" + } + } + ] +} diff --git a/compute/model-specs/stealth.json b/compute/model-specs/stealth.json new file mode 100644 index 0000000..a672513 --- /dev/null +++ b/compute/model-specs/stealth.json @@ -0,0 +1,46 @@ +{ + "description": "Stealth 系列模型规格。参数来源：OpenRouter Models API。这些是厂商匿名发布的临时别名，揭晓后用 spec.extra.modelOrigin 标注真身，条目本身保留——云端算力可能仍以旧别名下发该模型。", + "specs": [ + { + "id": "ox-alpha", + "displayName": "Ox Alpha (Historical)", + "family": "ox-alpha", + "match": { + "exact": [ + "ox-alpha", + "stealth/ox-alpha" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "image_understanding", + "video_understanding" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "Ox Alpha 历史匿名模型兼容记录；当前官方目录未列出。旧记录曾关联 GLM-5.3-Flash，但本轮未取得官方身份揭示证据，不下发未证实的原厂身份映射；历史参数不代表当前可用性", + "extra": { + "thinkingDefault": true, + "thinkingOnly": true + }, + "releasedAt": "2026-08-21" + } + } + ] +} diff --git a/compute/model-specs/tencent.json b/compute/model-specs/tencent.json new file mode 100644 index 0000000..6379d32 --- /dev/null +++ b/compute/model-specs/tencent.json @@ -0,0 +1,265 @@ +{ + "description": "腾讯混元系列模型规格。参数来源：config-center compute/providers/tencent.json。hunyuan-2.0-thinking 和 hunyuan-t1 为推理模型。", + "specs": [ + { + "id": "hy4-preview", + "displayName": "Tencent Hy4 Preview", + "family": "hy4-preview", + "match": { + "exact": [ + "hy4-preview", + "tencent/hy4-preview" + ] + }, + "spec": { + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "Tencent Hy4 Preview，通过 TokenHub 提供；既有 Hunyuan Provider 端点可用性尚未确认", + "contextWindow": 1048576, + "releasedAt": "2026-08-28" + } + }, + { + "id": "hy3", + "displayName": "混元 Hy3", + "family": "hy3", + "match": { + "exact": [ + "hy3" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 0.9, + "supportsReasoning": true, + "description": "腾讯混元 Hy3 正式版，295B MoE（21B active），支持可配置推理强度", + "releasedAt": "2026-07-06" + } + }, + { + "id": "hy3-preview", + "displayName": "混元 Hy3 Preview", + "family": "hy3", + "match": { + "exact": [ + "hy3-preview" + ], + "patterns": [ + "hy3-preview*" + ] + }, + "spec": { + "contextWindow": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "腾讯混元 Hy3 预览版，高效 MoE 模型" + } + }, + { + "id": "hunyuan-2.0-thinking-20251109", + "displayName": "混元 2.0 Think", + "family": "hunyuan-2", + "match": { + "exact": [ + "hunyuan-2.0-thinking-20251109" + ], + "patterns": [ + "hunyuan-2.0-thinking*" + ] + }, + "spec": { + "contextWindow": 196608, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "腾讯混元最新推理模型，MoE架构406B总参数，深度思考" + } + }, + { + "id": "hunyuan-2.0-instruct-20251111", + "displayName": "混元 2.0 Instruct", + "family": "hunyuan-2", + "match": { + "exact": [ + "hunyuan-2.0-instruct-20251111" + ], + "patterns": [ + "hunyuan-2.0-instruct*" + ] + }, + "spec": { + "contextWindow": 147456, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "腾讯混元 2.0 指令模型，支持联网搜索和 Function Calling" + } + }, + { + "id": "hunyuan-turbos-latest", + "displayName": "混元 Turbo S", + "family": "hunyuan-turbo", + "match": { + "exact": [ + "hunyuan-turbos-latest" + ], + "patterns": [ + "hunyuan-turbos*" + ] + }, + "spec": { + "contextWindow": 32768, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "fast", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "腾讯混元高速模型，支持推理和工具调用" + } + }, + { + "id": "hunyuan-t1-latest", + "displayName": "混元 T1", + "family": "hunyuan-t1", + "match": { + "exact": [ + "hunyuan-t1-latest" + ], + "patterns": [ + "hunyuan-t1*" + ] + }, + "spec": { + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "腾讯混元 T1 深度推理模型，256K 上下文，专攻数学/代码/复杂推理" + } + }, + { + "id": "hunyuan-t1-vision", + "displayName": "混元 T1 Vision", + "family": "hunyuan-t1", + "match": { + "exact": [ + "hunyuan-t1-vision" + ] + }, + "spec": { + "contextWindow": 131072, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "vision", + "image_understanding" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "腾讯混元 T1 视觉推理模型，支持图片理解 + 深度推理" + } + }, + { + "id": "hunyuan-turbos-vision", + "displayName": "混元 Turbo S Vision", + "family": "hunyuan-turbo", + "match": { + "exact": [ + "hunyuan-turbos-vision" + ] + }, + "spec": { + "contextWindow": 32768, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "image_understanding", + "fast", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "腾讯混元 Turbo S 视觉模型，支持图片理解，高速推理" + } + } + ] +} diff --git a/compute/model-specs/volcengine.json b/compute/model-specs/volcengine.json new file mode 100644 index 0000000..48604a7 --- /dev/null +++ b/compute/model-specs/volcengine.json @@ -0,0 +1,808 @@ +{ + "description": "火山引擎豆包系列模型规格。参数来源：config-center compute/providers/volcengine.json。包含豆包 Seed 系列对话、DeepSeek 托管版、Embedding、语音克隆、实时语音、同传和翻译服务。doubao-seed-1.6-thinking 为推理模型。", + "specs": [ + { + "id": "doubao-seed-evolving", + "displayName": "Doubao Seed Evolving", + "family": "doubao-seed-evolving", + "match": { + "exact": [ + "doubao-seed-evolving" + ] + }, + "spec": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "Seed 系列持续迭代模型，通过固定 ID 获取最新版本，2026-09-09 更新" + } + }, + { + "id": "doubao-seed-2.1-pro", + "displayName": "豆包 Seed 2.1 Pro", + "family": "doubao-seed-2.1-pro", + "match": { + "exact": [ + "doubao-seed-2.1-pro" + ] + }, + "spec": { + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "火山方舟豆包 Seed 2.1 Pro，生产级 Coding、Agent 与多模态旗舰模型", + "contextWindow": 256000, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "long_context", + "tool_use", + "agent", + "deep_thinking" + ], + "supportsReasoning": true + } + }, + { + "id": "doubao-seed-2.0-pro", + "displayName": "豆包 2.0 Pro", + "family": "doubao-seed-2", + "match": { + "exact": [ + "doubao-seed-2.0-pro" + ], + "patterns": [ + "doubao-seed-2.0-pro*" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 Seed 2.0 旗舰模型，256K 上下文，支持复杂推理、Agent 和多模态" + } + }, + { + "id": "doubao-seed-2.0-lite", + "displayName": "豆包 2.0 Lite", + "family": "doubao-seed-2", + "match": { + "exact": [ + "doubao-seed-2.0-lite" + ], + "patterns": [ + "doubao-seed-2.0-lite*" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "audio_understanding", + "video_understanding" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "豆包 Seed 2.0 全模态均衡模型，支持视频/图像/音频/文本统一理解" + } + }, + { + "id": "doubao-seed-2.0-mini", + "displayName": "豆包 2.0 Mini", + "family": "doubao-seed-2", + "match": { + "exact": [ + "doubao-seed-2.0-mini" + ], + "patterns": [ + "doubao-seed-2.0-mini*" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "audio_understanding", + "video_understanding", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "豆包 Seed 2.0 全模态轻量模型，低延迟高并发，支持全模态理解" + } + }, + { + "id": "doubao-seed-2.0-code", + "displayName": "豆包 2.0 Code", + "family": "doubao-seed-2", + "match": { + "exact": [ + "doubao-seed-2.0-code" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "long_context", + "tool_use", + "vision", + "deep_thinking" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "豆包 Seed 2.0 编程模型，面向 Agentic Coding 场景优化" + } + }, + { + "id": "doubao-seed-1.8", + "displayName": "豆包 Seed-1.8", + "family": "doubao-seed-1", + "match": { + "exact": [ + "doubao-seed-1.8" + ], + "patterns": [ + "doubao-seed-1.8*" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.8 旗舰模型，256K 上下文，支持深度思考和多模态" + } + }, + { + "id": "doubao-seed-1.6", + "displayName": "豆包 Seed-1.6", + "family": "doubao-seed-1", + "match": { + "exact": [ + "doubao-seed-1.6" + ], + "patterns": [ + "doubao-seed-1.6" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 旗舰模型，256K 上下文" + } + }, + { + "id": "doubao-seed-1.6-thinking", + "displayName": "豆包 Seed-1.6 思考版", + "family": "doubao-seed-1", + "match": { + "exact": [ + "doubao-seed-1.6-thinking" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "火山引擎豆包 1.6 深度思考模型，256K 上下文" + } + }, + { + "id": "doubao-seed-1.6-flash", + "displayName": "豆包 Seed-1.6 Flash", + "family": "doubao-seed-1", + "match": { + "exact": [ + "doubao-seed-1.6-flash" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "chat", + "code", + "fast", + "long_context", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 快速版，成本较前代下降 63%" + } + }, + { + "id": "doubao-seed-1.6-lite", + "displayName": "豆包 Seed-1.6 Lite", + "family": "doubao-seed-1", + "match": { + "exact": [ + "doubao-seed-1.6-lite" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "code", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 轻量版，极致性价比" + } + }, + { + "id": "doubao-seed-1.6-vision", + "displayName": "豆包 Seed-1.6 视觉版", + "family": "doubao-seed-1", + "match": { + "exact": [ + "doubao-seed-1.6-vision" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "gui_agent", + "long_context" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 视觉模型，256K 上下文，支持视频理解" + } + }, + { + "id": "doubao-seed-code", + "displayName": "豆包 Seed-Code", + "family": "doubao-seed-code", + "match": { + "exact": [ + "doubao-seed-code" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 32768, + "capabilities": [ + "code", + "reasoning", + "long_context", + "deep_thinking" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包编程专用模型，256K 上下文" + } + }, + { + "id": "deepseek-v3.2", + "displayName": "DeepSeek V3.2 (火山引擎)", + "family": "deepseek-v3", + "match": { + "exact": [ + "deepseek-v3.2" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎托管的 DeepSeek V3.2，128K 上下文" + } + }, + { + "id": "deepseek-r1", + "displayName": "DeepSeek R1 (火山引擎)", + "family": "deepseek-r1", + "match": { + "exact": [ + "deepseek-r1" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 65536, + "capabilities": [ + "reasoning", + "math", + "code", + "science", + "deep_thinking" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "火山引擎托管的 DeepSeek R1 推理模型，128K 上下文" + } + }, + { + "id": "doubao-embedding", + "displayName": "豆包 Embedding", + "family": "doubao-embedding", + "match": { + "exact": [ + "doubao-embedding" + ] + }, + "spec": { + "contextWindow": 4096, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "chinese_optimized" + ], + "serviceType": [ + "embedding" + ], + "description": "火山引擎豆包标准向量模型，默认 2048 维度，可选 512/1024/2048" + } + }, + { + "id": "doubao-embedding-large", + "displayName": "豆包 Embedding Large", + "family": "doubao-embedding", + "match": { + "exact": [ + "doubao-embedding-large" + ] + }, + "spec": { + "contextWindow": 4096, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "chinese_optimized" + ], + "serviceType": [ + "embedding" + ], + "description": "火山引擎豆包大号向量模型，默认 2048 维度，可选 512/1024/2048/4096" + } + }, + { + "id": "volc-mega-tts-clone", + "displayName": "火山声音复刻", + "family": "volc-tts", + "match": { + "exact": [ + "volc-mega-tts-clone" + ] + }, + "spec": { + "capabilities": [ + "voice_clone", + "chinese_optimized", + "few_shot" + ], + "serviceType": [ + "voice_clone" + ], + "description": "火山引擎声音复刻，支持极少样本克隆" + } + }, + { + "id": "volc-realtime-voice", + "displayName": "火山实时语音", + "family": "volc-voice", + "match": { + "exact": [ + "volc-realtime-voice" + ] + }, + "spec": { + "capabilities": [ + "realtime", + "speech_to_speech", + "low_latency", + "chinese_optimized" + ], + "serviceType": [ + "realtime_voice" + ], + "description": "火山引擎实时语音交互" + } + }, + { + "id": "volc-simultaneous", + "displayName": "火山同声传译", + "family": "volc-interpret", + "match": { + "exact": [ + "volc-simultaneous" + ] + }, + "spec": { + "capabilities": [ + "simultaneous_interpretation", + "chinese_optimized", + "low_latency" + ], + "serviceType": [ + "simultaneous_interpret" + ], + "description": "火山引擎同声传译，支持中英双向" + } + }, + { + "id": "volc-translation", + "displayName": "火山翻译", + "family": "volc-translation", + "match": { + "exact": [ + "volc-translation" + ] + }, + "spec": { + "capabilities": [ + "translation", + "multilingual", + "chinese_optimized" + ], + "serviceType": [ + "translation" + ], + "description": "火山引擎机器翻译，支持多语种互译" + } + }, + { + "id": "doubao-seedance-2.0", + "displayName": "Seedance 2.0", + "family": "doubao-seedance-2", + "match": { + "exact": [ + "doubao-seedance-2-0-260128", + "doubao-seedance-2.0" + ], + "patterns": [ + "doubao-seedance-2-0*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "video_editing", + "video_extension", + "audio_generation", + "first_last_frame" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 2.0 豆包最强视频生成模型，4-15秒 2K 视频，原生同步音频，支持最多9张图片+3个视频+3个音频参考输入，~$0.14/秒" + }, + "routing": { + "tier": "flagship", + "routingPriority": 120, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "doubao-seedance-2.0-fast", + "displayName": "Seedance 2.0 Fast", + "family": "doubao-seedance-2", + "match": { + "exact": [ + "doubao-seedance-2-0-fast-260128", + "doubao-seedance-2.0-fast" + ], + "patterns": [ + "doubao-seedance-2-0-fast*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "video_editing", + "video_extension", + "audio_generation", + "first_last_frame", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 2.0 快速版，音画同生，多模态生视频+编辑+延长" + }, + "routing": { + "tier": "balanced", + "routingPriority": 119, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "doubao-seedance-2.0-mini", + "displayName": "Seedance 2.0 Mini", + "family": "doubao-seedance-2", + "match": { + "exact": [ + "doubao-seedance-2-0-mini-260615", + "doubao-seedance-2.0-mini" + ], + "patterns": [ + "doubao-seedance-2-0-mini*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "video_editing", + "video_extension", + "audio_generation", + "first_last_frame", + "fast", + "cost_effective" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 2.0 轻量版，音画同生，高性价比" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 118, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "doubao-seedance-1.5-pro", + "displayName": "Seedance 1.5 Pro", + "family": "doubao-seedance-1", + "match": { + "exact": [ + "doubao-seedance-1-5-pro-251215", + "doubao-seedance-1.5-pro" + ], + "patterns": [ + "doubao-seedance-1-5*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "first_last_frame", + "audio_generation" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 1.5 Pro 音画同生，12秒视频，多参考图输入（即将下线）" + } + }, + { + "id": "doubao-seedance-1.0-pro", + "displayName": "Seedance 1.0 Pro", + "family": "doubao-seedance-1", + "match": { + "exact": [ + "doubao-seedance-1-0-pro-250528", + "doubao-seedance-1.0-pro" + ], + "patterns": [ + "doubao-seedance-1-0-pro*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "first_last_frame" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 1.0 Pro 文生视频/图生视频，支持首尾帧控制" + } + }, + { + "id": "doubao-seedance-1.0-pro-fast", + "displayName": "Seedance 1.0 Pro Fast", + "family": "doubao-seedance-1", + "match": { + "exact": [ + "doubao-seedance-1-0-pro-fast-251015", + "doubao-seedance-1.0-pro-fast" + ], + "patterns": [ + "doubao-seedance-1-0-pro-fast*" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 1.0 Pro 快速版，图生视频-首帧，5秒 720P 仅需 10 秒" + } + }, + { + "id": "doubao-seed-2.1-turbo", + "displayName": "Doubao Seed 2.1 Turbo", + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + } + ] +} diff --git a/compute/model-specs/wan.json b/compute/model-specs/wan.json new file mode 100644 index 0000000..f78716c --- /dev/null +++ b/compute/model-specs/wan.json @@ -0,0 +1,106 @@ +{ + "description": "通义万相（Wan）多模态生成模型规格。Wan 3.0 视频模型通过阿里云百炼提供，统一支持文生视频、首帧/首尾帧图生视频和多类型参考生视频。", + "specs": [ + { + "id": "wan3.0-video-prime", + "displayName": "通义万相 3.0 视频生成高速版", + "family": "wan3-video", + "match": { + "exact": [ + "wan3.0-video-prime" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "high_quality", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "万相 3.0 全能参考视频生成高速版，能力对齐标准版并显著提升端到端生成速度", + "releasedAt": "2026-08-06", + "extra": { + "maxVideoDuration": 30, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ] + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 109, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "wan3.0-video", + "displayName": "通义万相 3.0 视频生成", + "family": "wan3-video", + "match": { + "exact": [ + "wan3.0-video" + ] + }, + "spec": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "万相 3.0 全能参考视频生成标准版，统一支持文生、首帧/首尾帧图生及多类型参考生视频", + "releasedAt": "2026-08-06", + "extra": { + "maxVideoDuration": 30, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ] + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 110, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + } + ] +} diff --git a/compute/model-specs/xai.json b/compute/model-specs/xai.json new file mode 100644 index 0000000..413948d --- /dev/null +++ b/compute/model-specs/xai.json @@ -0,0 +1,155 @@ +{ + "description": "xAI Grok 系列模型规格。参数来源：config-center compute/providers/xai.json。Grok 网关命名变体较多（含日期戳/-fast/-reasoning 后缀），主要依赖 family=grok-4 兜底匹配。", + "specs": [ + { + "id": "grok-4.7", + "displayName": "Grok 4.7", + "family": "grok-4.7", + "match": { + "exact": [ + "grok-4.7" + ] + }, + "spec": { + "contextWindow": 500000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "xAI Grok 4.7，500K 上下文，文本与图像输入，推理努力等级 low 到 xhigh" + }, + "routing": { + "tier": "flagship", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + } + }, + { + "id": "grok-4.5", + "displayName": "Grok 4.5", + "family": "grok-4.5", + "match": { + "exact": [ + "grok-4.5" + ] + }, + "spec": { + "contextWindow": 500000, + "maxOutputTokens": 450000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "xAI Grok 4.5，面向编程、知识工作和 STEM 的旗舰推理模型", + "extra": { + "thinkingOnly": true, + "thinkingDefault": true + }, + "releasedAt": "2026-07-08" + } + }, + { + "id": "grok-4-3", + "displayName": "Grok 4.3", + "family": "grok-4", + "match": { + "exact": [ + "grok-4.3" + ], + "patterns": [ + "grok-4.3*" + ] + }, + "spec": { + "contextWindow": 1000000, + "capabilities": [ + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "xAI Grok 4.3 推理模型，100 万上下文" + } + }, + { + "id": "grok-4-1-fast-reasoning", + "displayName": "Grok 4.1 Fast", + "family": "grok-4", + "spec": { + "contextWindow": 2000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "fast", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "grok-4.20-0309-reasoning", + "displayName": "Grok 4.20", + "family": "grok-4", + "spec": { + "contextWindow": 2000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + } + ] +} diff --git a/compute/model-specs/xiaomi.json b/compute/model-specs/xiaomi.json new file mode 100644 index 0000000..7aa5c04 --- /dev/null +++ b/compute/model-specs/xiaomi.json @@ -0,0 +1,556 @@ +{ + "description": "小米 MiMo 系列模型规格。参数来源：https://mimo.mi.com/docs/zh-CN/quick-start/summary/model。注意 v2-pro/v2-omni 自 2026-06-01 起转发到 v2.5 系列，2026-06-30 正式停用，规格库仍保留供存量 NewAPI 渠道正确展示参数。", + "specs": [ + { + "id": "mimo-v2.6-pro", + "displayName": "MiMo-V2.6-Pro", + "family": "mimo-v2.6-pro", + "match": { + "exact": [ + "mimo-v2.6-pro" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "小米全模态旗舰推理模型，适合复杂项目与长程智能体任务", + "extra": { + "thinkingDefault": true + }, + "releasedAt": "2026-09-22" + }, + "routing": { + "tier": "flagship", + "routingPriority": 28, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "mimo-v2.6-flash", + "displayName": "MiMo-V2.6-Flash", + "family": "mimo-v2.6-flash", + "match": { + "exact": [ + "mimo-v2.6-flash" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "小米全模态高效推理模型，适合高频调用与规模化任务", + "extra": { + "thinkingDefault": true + }, + "releasedAt": "2026-09-22" + }, + "routing": { + "tier": "balanced", + "routingPriority": 28, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "mimo-v2.6-pro-ultraspeed", + "displayName": "MiMo-V2.6-Pro-UltraSpeed", + "family": "mimo-v2.6-pro-ultraspeed", + "match": { + "exact": [ + "mimo-v2.6-pro-ultraspeed" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "MiMo-V2.6-Pro 的超高速服务，需向小米申请定制配额", + "extra": { + "thinkingDefault": true + }, + "releasedAt": "2026-09-22" + } + }, + { + "id": "mimo-x-flash-preview", + "displayName": "MiMo-X-Flash Preview", + "family": "mimo-x-flash-preview", + "match": { + "exact": [ + "mimo-x-flash-preview" + ] + }, + "spec": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "小米 MiMo-X Flash 内测模型，1M 上下文，支持图片识别" + } + }, + { + "id": "mimo-x-pro-preview", + "displayName": "MiMo-X-Pro Preview", + "family": "mimo-x-pro-preview", + "match": { + "exact": [ + "mimo-x-pro-preview" + ] + }, + "spec": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "小米 MiMo-X Pro 内测模型，1M 上下文，支持图片识别" + } + }, + { + "id": "mimo-v2.5-pro", + "displayName": "MiMo V2.5 Pro", + "family": "mimo-v2.5-pro", + "match": { + "patterns": [ + "mimo-v2.5-pro*", + "mimo-v2-5-pro*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo 旗舰推理模型，支持文本/工具调用" + }, + "routing": { + "tier": "flagship", + "routingPriority": 50, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + } + }, + { + "id": "mimo-v2.5", + "displayName": "MiMo V2.5", + "family": "mimo-v2.5", + "match": { + "exact": [ + "mimo-v2.5" + ], + "patterns": [ + "mimo-v2.5-chat*", + "mimo-v2-5*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo 全模态模型，支持图像/音频/视频理解" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "low" + } + } + }, + { + "id": "mimo-v2-pro", + "displayName": "MiMo V2 Pro（已停用→转发至 V2.5 Pro）", + "family": "mimo-v2-pro", + "match": { + "patterns": [ + "mimo-v2-pro*" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo V2 Pro（自 2026-06-01 转发至 V2.5 Pro）", + "retiredAt": "2026-06-30" + } + }, + { + "id": "mimo-v2-omni", + "displayName": "MiMo V2 Omni（已停用→转发至 V2.5）", + "family": "mimo-v2-omni", + "match": { + "patterns": [ + "mimo-v2-omni*" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo V2 全模态（自 2026-06-01 转发至 V2.5）", + "retiredAt": "2026-06-30" + } + }, + { + "id": "mimo-v2-flash", + "displayName": "MiMo V2 Flash", + "family": "mimo-v2-flash", + "match": { + "patterns": [ + "mimo-v2-flash*" + ] + }, + "spec": { + "contextWindow": 256000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo 快速版，适合高频低延迟任务" + } + }, + { + "id": "mimo-v2.5-tts", + "displayName": "MiMo V2.5 TTS", + "family": "mimo-v2.5-tts", + "match": { + "exact": [ + "mimo-v2.5-tts" + ] + }, + "spec": { + "contextWindow": 8192, + "maxOutputTokens": 8192, + "capabilities": [ + "tts", + "multilingual", + "style_control" + ], + "serviceType": [ + "tts" + ], + "description": "小米 MiMo 语音合成，支持多语言与风格控制", + "extra": { + "modelOrigin": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-tts", + "family": "mimo-tts" + }, + "speech": { + "protocol": "xiaomi-mimo-chat-tts", + "protocolVersion": "1", + "modes": [ + "batch", + "streaming" + ], + "outputFormats": [ + "wav", + "mp3", + "pcm16le" + ], + "sampleRates": [ + 24000 + ], + "defaultVoice": "mimo_default", + "voices": [ + { "id": "mimo_default", "label": "MiMo 默认" }, + { "id": "冰糖", "label": "冰糖" }, + { "id": "茉莉", "label": "茉莉" }, + { "id": "苏打", "label": "苏打" }, + { "id": "白桦", "label": "白桦" }, + { "id": "Mia", "label": "Mia" }, + { "id": "Chloe", "label": "Chloe" }, + { "id": "Milo", "label": "Milo" }, + { "id": "Dean", "label": "Dean" } + ] + } + }, + "releasedAt": "2026-04-23" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 114, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "mimo-v2.5-asr", + "displayName": "MiMo V2.5 ASR", + "family": "mimo-v2.5-asr", + "match": { + "exact": [ + "mimo-v2.5-asr" + ] + }, + "spec": { + "contextWindow": 8192, + "maxOutputTokens": 2048, + "capabilities": [ + "asr", + "multilingual" + ], + "serviceType": [ + "asr" + ], + "description": "小米 MiMo 语音识别模型，支持中英文语言提示与整段音频识别", + "extra": { + "modelOrigin": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-asr", + "family": "mimo-asr" + }, + "speech": { + "protocol": "xiaomi-mimo-chat-asr", + "protocolVersion": "1", + "modes": [ + "batch" + ], + "inputFormats": [ + "wav", + "mp3" + ], + "languageHints": true + } + }, + "releasedAt": "2026-06-02" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 123, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + } + }, + { + "id": "mimo-v2.5-tts-voiceclone", + "displayName": "MiMo V2.5 TTS Voice Clone", + "family": "mimo-v2.5-tts", + "match": { + "exact": [ + "mimo-v2.5-tts-voiceclone" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "tts", + "voice_clone", + "multilingual" + ], + "serviceType": [ + "tts", + "voice_clone" + ], + "description": "小米 MiMo 声音复刻合成" + } + }, + { + "id": "mimo-v2.5-tts-voicedesign", + "displayName": "MiMo V2.5 TTS Voice Design", + "family": "mimo-v2.5-tts", + "match": { + "exact": [ + "mimo-v2.5-tts-voicedesign" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "tts", + "voice_design", + "multilingual" + ], + "serviceType": [ + "tts" + ], + "description": "小米 MiMo 自定义音色设计合成" + } + }, + { + "id": "mimo-v2-tts", + "displayName": "MiMo V2 TTS", + "family": "mimo-v2-tts", + "match": { + "exact": [ + "mimo-v2-tts" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "tts", + "multilingual" + ], + "serviceType": [ + "tts" + ], + "description": "小米 MiMo V2 语音合成（旧版）" + } + } + ] +} diff --git a/compute/model-specs/xunfei.json b/compute/model-specs/xunfei.json new file mode 100644 index 0000000..b5470c1 --- /dev/null +++ b/compute/model-specs/xunfei.json @@ -0,0 +1,61 @@ +{ + "description": "讯飞星火系列模型规格。参数来源：config-center compute/providers/xunfei.json。spark-x 为推理模型（serviceType: reasoning）。", + "specs": [ + { + "id": "spark-x", + "displayName": "讯飞星火 X2", + "family": "spark-x", + "match": { + "exact": [ + "spark-x" + ] + }, + "spec": { + "contextWindow": 65536, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "math", + "code", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": 1.2, + "supportsReasoning": true, + "description": "讯飞星火 X2，/x2 接入合同：输入 65536 / 输出 131072；spark-x 同名在 /v2 上指向 X1.5，不能复用本输出限制" + } + }, + { + "id": "4.0Ultra", + "displayName": "讯飞星火 4.0 Ultra", + "family": "spark-4", + "match": { + "exact": [ + "4.0Ultra" + ], + "patterns": [ + "4.0Ultra*" + ] + }, + "spec": { + "contextWindow": 32768, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "讯飞星火最强通用模型，支持推理、代码和工具调用" + } + } + ] +} diff --git a/compute/model-specs/zhipu.json b/compute/model-specs/zhipu.json new file mode 100644 index 0000000..4351bd9 --- /dev/null +++ b/compute/model-specs/zhipu.json @@ -0,0 +1,462 @@ +{ + "description": "智谱 GLM 系列模型规格。参数来源：config-center compute/providers/zhipu.json。glm-5 与 glm-5.1/glm-5-turbo/glm-5v-turbo 前缀相近，故各自仅用 exact 主键匹配，不用宽 pattern 以防误吞。", + "specs": [ + { + "id": "glm-5.3-flashx", + "displayName": "GLM-5.3-FlashX", + "family": "glm-5.3-flashx", + "match": { + "exact": [ + "glm-5.3-flashx", + "z-ai/glm-5.3-flashx" + ] + }, + "spec": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "video_understanding", + "tool_use", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "Z.ai GLM-5.3-FlashX，高速多模态推理模型，100 万上下文，始终开启推理", + "extra": { + "adaptiveThinking": true, + "thinkingOnly": true + } + }, + "routing": { + "tier": "balanced", + "routingPriority": 49, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + } + }, + { + "id": "glm-5.3", + "displayName": "GLM-5.3", + "family": "glm-5.3", + "match": { + "exact": [ + "glm-5.3", + "z-ai/glm-5.3" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "extra": { + "adaptiveThinking": true, + "thinkingDefault": true, + "thinkingOnly": true + }, + "releasedAt": "2026-08-16" + }, + "routing": { + "tier": "flagship", + "routingPriority": 50, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + } + }, + { + "id": "glm-5.3-flash", + "displayName": "GLM-5.3-Flash", + "family": "glm-5.3-flash", + "match": { + "exact": [ + "glm-5.3-flash", + "z-ai/glm-5.3-flash" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "image_understanding", + "video_understanding" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "智谱 GLM-5.3-Flash 原生多模态模型，混合稀疏与线性注意力，面向高效编码与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "extra": { + "adaptiveThinking": true, + "thinkingDefault": true, + "thinkingOnly": true + }, + "releasedAt": "2026-08-26" + }, + "routing": { + "tier": "balanced", + "routingPriority": 48, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + } + }, + { + "id": "glm-5.2", + "displayName": "GLM-5.2", + "family": "glm-5.2", + "match": { + "exact": [ + "glm-5.2" + ], + "patterns": [ + "glm-5.2*" + ] + }, + "spec": { + "contextWindow": 1048576, + "maxOutputTokens": 32768, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "智谱 GLM-5.2 大推理模型，100 万上下文" + }, + "routing": { + "tier": "flagship", + "routingPriority": 45, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + } + }, + { + "id": "glm-5.1", + "displayName": "GLM-5.1", + "family": "glm-5.1", + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 65, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + } + }, + { + "id": "glm-5", + "displayName": "GLM-5", + "family": "glm-5", + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "glm-5-turbo", + "displayName": "GLM-5 Turbo", + "family": "glm-5-turbo", + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "glm-4.7", + "displayName": "GLM-4.7", + "family": "glm-4.7", + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "glm-4.6", + "displayName": "GLM-4.6", + "family": "glm-4.6", + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + } + }, + { + "id": "glm-4.7-thinking", + "displayName": "GLM-4.7 Thinking", + "family": "glm-4.7", + "match": { + "exact": [ + "glm-4.7-thinking" + ] + }, + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "智谱GLM-4.7深度思考模式，交错式/保留式/轮级思考" + } + }, + { + "id": "glm-5v-turbo", + "displayName": "GLM-5V-Turbo", + "family": "glm-5v", + "match": { + "exact": [ + "glm-5v-turbo" + ] + }, + "spec": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "file_understanding", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "智谱首个多模态 Coding 基座模型，支持视频、图像、文本和文件输入" + } + }, + { + "id": "glm-4.6v", + "displayName": "GLM-4.6V", + "family": "glm-4.6", + "match": { + "exact": [ + "glm-4.6v" + ] + }, + "spec": { + "contextWindow": 128000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "long_context", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1, + "description": "智谱GLM-4.6V多模态版，106B/12B MoE，支持图像视频理解" + } + }, + { + "id": "embedding-3", + "displayName": "智谱 embedding-3", + "family": "zhipu-embedding", + "match": { + "exact": [ + "embedding-3" + ] + }, + "spec": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "custom_dimensions" + ], + "serviceType": [ + "embedding" + ], + "description": "智谱嵌入模型v3，支持自定义维度；单条输入最多 3072 tokens" + } + } + ] +} diff --git a/compute/pricing.json b/compute/pricing.json new file mode 100644 index 0000000..b54629a --- /dev/null +++ b/compute/pricing.json @@ -0,0 +1,4 @@ +{ + "markupRatio": 1.5, + "usdToCny": 7.2 +} diff --git a/compute/providers/_index.json b/compute/providers/_index.json new file mode 100644 index 0000000..3d85aa3 --- /dev/null +++ b/compute/providers/_index.json @@ -0,0 +1,34 @@ +{ + "description": "Provider 加载顺序索引", + "order": [ + "openai", + "openai-codex", + "anthropic", + "anthropic-claude", + "github-copilot", + "deepseek", + "dashscope", + "volcengine", + "moonshot", + "zhipu", + "zhipu-embedding", + "baichuan", + "minimax", + "xiaomi", + "baidu", + "tencent", + "xunfei", + "xunfei-x2", + "siliconflow", + "local-whisper", + "ollama", + "google", + "mistral", + "xai", + "cohere", + "kling", + "stability", + "openrouter", + "perplexity" + ] +} diff --git a/compute/providers/anthropic-claude.json b/compute/providers/anthropic-claude.json new file mode 100644 index 0000000..1437919 --- /dev/null +++ b/compute/providers/anthropic-claude.json @@ -0,0 +1,239 @@ +{ + "id": "provider-anthropic-claude-plan-001", + "provider": "anthropic-claude", + "brandGroup": "anthropic", + "label": "Claude 订阅 (Claude Code)", + "baseUrl": "https://api.anthropic.com", + "apiFormat": "anthropic-messages", + "apiKeyRef": "anthropic-claude-oauth-token", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "accessMode": "coding-plan", + "credentialSource": "claude-oauth", + "requiredClientVersion": "10.0.83", + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://claude.ai/settings/usage" + } + }, + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "claude-opus-5-5", + "displayName": "Claude Opus 5.5", + "serviceType": [ + "chat" + ], + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "description": "Claude 订阅额度计费，长程智能体编程与知识工作模型，1M 上下文", + "extra": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "thinkingOnly": true, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "source": "preset" + }, + { + "modelName": "claude-fable-5-1", + "displayName": "Claude Fable 5.1", + "serviceType": [ + "chat" + ], + "description": "Claude 订阅额度计费，面向高难推理和长程智能体任务，1M 上下文，始终开启自适应思考", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "extra": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingOnly": true, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "source": "preset" + }, + { + "modelName": "claude-opus-5", + "displayName": "Claude Opus 5", + "serviceType": [ + "chat" + ], + "description": "Claude 订阅额度计费，Opus 5 系列模型，复杂智能体编码与长程任务首选，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "extra": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "source": "preset" + }, + { + "modelName": "claude-sonnet-5-5", + "displayName": "Claude Sonnet 5.5", + "serviceType": [ + "chat" + ], + "description": "Claude Sonnet 5.5，速度与智能平衡，1M 上下文，默认 high effort，自适应思考", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "extra": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + }, + "source": "preset" + }, + { + "modelName": "claude-sonnet-5", + "displayName": "Claude Sonnet 5", + "serviceType": [ + "chat" + ], + "description": "Claude 订阅额度计费，速度与智能平衡模型，支持长上下文智能体任务", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "extra": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "source": "preset" + } + ], + "tombstones": [ + "claude-opus-4-8", + "claude-fable-5", + "claude-haiku-4-5", + "claude-haiku-4-5-20251001" + ] +} diff --git a/compute/providers/anthropic.json b/compute/providers/anthropic.json new file mode 100644 index 0000000..a4f8621 --- /dev/null +++ b/compute/providers/anthropic.json @@ -0,0 +1,311 @@ +{ + "id": "provider-anthropic-001", + "provider": "anthropic", + "label": "Anthropic Claude", + "baseUrl": "https://api.anthropic.com", + "apiFormat": "anthropic-messages", + "apiKeyRef": "anthropic", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "computer_use" + ], + "models": [ + { + "modelName": "claude-opus-5-5", + "displayName": "Claude Opus 5.5", + "serviceType": [ + "chat" + ], + "description": "Anthropic 面向长程智能体编程与知识工作的 Opus 模型，1M 上下文，始终开启自适应思考", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 4, + "outputPrice": 20, + "extra": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 2, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search_20260318" + }, + "cachePricing": { + "write5m": 5, + "write1h": 8, + "read": 0.2 + }, + "promptCacheMinTokens": 512, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true, + "thinkingOnly": true, + "pricingNotes": "Prices are per 1M tokens. Full 1M context is billed at standard pricing." + } + }, + { + "modelName": "claude-fable-5-1", + "displayName": "Claude Fable 5.1", + "serviceType": [ + "chat" + ], + "description": "面向高难推理和长程智能体任务，1M 上下文，始终开启自适应思考", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 10, + "outputPrice": 50, + "extra": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 1, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search_20260318" + }, + "cachePricing": { + "write5m": 12.5, + "write1h": 20, + "read": 0.25 + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true, + "pricingNotes": "Prices are per 1M tokens. Full 1M context is billed at standard pricing.", + "thinkingOnly": true + } + }, + { + "modelName": "claude-opus-5", + "displayName": "Claude Opus 5", + "serviceType": [ + "chat" + ], + "description": "Anthropic Opus 5 系列模型，复杂智能体编码与长程任务能力显著强于 Opus 4.8，价格不变，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 5, + "outputPrice": 25, + "extra": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 2, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search_20260318" + }, + "cachePricing": { + "write5m": 6.25, + "write1h": 10, + "read": 0.5 + }, + "promptCacheMinTokens": 512, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "samplingParametersDeprecated": true, + "pricingNotes": "Prices are per 1M tokens. Full 1M context is billed at standard pricing." + } + }, + { + "modelName": "claude-sonnet-5-5", + "displayName": "Claude Sonnet 5.5", + "serviceType": [ + "chat", + "computer_use" + ], + "description": "Claude Sonnet 5.5，速度与智能平衡，1M 上下文，默认 high effort，自适应思考", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "agent", + "long_context" + ], + "inputPrice": 2, + "outputPrice": 10, + "extra": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 3, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search_20260318" + }, + "cachePricing": { + "write5m": 2.5, + "write1h": 4, + "read": 0.2 + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true, + "promptCacheMinTokens": 512 + } + }, + { + "modelName": "claude-sonnet-5", + "displayName": "Claude Sonnet 5", + "serviceType": [ + "chat", + "computer_use" + ], + "description": "Anthropic Sonnet 5，速度与智能平衡模型，支持电脑使用和长上下文智能体任务", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "agent", + "long_context" + ], + "inputPrice": 2, + "outputPrice": 10, + "extra": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 3, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search_20260318" + }, + "cachePricing": { + "write5m": 2.5, + "write1h": 4, + "read": 0.2 + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "samplingParametersDeprecated": true + } + } + ], + "tombstones": [ + "claude-opus-4-8", + "claude-opus-4-7", + "claude-sonnet-4-6", + "claude-fable-5", + "claude-haiku-4-5", + "claude-haiku-4-5-20251001" + ] +} diff --git a/compute/providers/baichuan.json b/compute/providers/baichuan.json new file mode 100644 index 0000000..ddc4aa9 --- /dev/null +++ b/compute/providers/baichuan.json @@ -0,0 +1,107 @@ +{ + "id": "provider-baichuan-001", + "provider": "baichuan", + "label": "百川 Baichuan", + "baseUrl": "https://api.baichuan-ai.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "baichuan", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat" + ], + "models": [ + { + "modelName": "Baichuan-M3-Plus", + "displayName": "百川 M3-Plus", + "serviceType": [ + "chat" + ], + "description": "百川智能M3-Plus医疗增强模型，235B参数，幻觉率仅2.6%，API价格较前代下降70%", + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "long_context", + "medical" + ], + "inputPrice": 5, + "outputPrice": 9, + "defaultTemperature": 0.3, + "defaultTopP": 0.85, + "extra": {} + }, + { + "modelName": "Baichuan-M3", + "displayName": "百川 M3", + "serviceType": [ + "chat" + ], + "description": "百川智能M3新一代开源医疗增强模型，235B参数", + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "medical" + ], + "inputPrice": 10, + "outputPrice": 30, + "defaultTemperature": 0.3, + "defaultTopP": 0.85, + "extra": {} + }, + { + "modelName": "Baichuan-M2-Plus", + "displayName": "百川 M2-Plus", + "serviceType": [ + "chat" + ], + "description": "百川智能M2-Plus旗舰模型，最强推理和创作能力", + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context" + ], + "inputPrice": 10, + "outputPrice": 30, + "defaultTemperature": 0.3, + "defaultTopP": 0.85, + "extra": {} + }, + { + "modelName": "Baichuan-M2", + "displayName": "百川 M2", + "serviceType": [ + "chat" + ], + "description": "百川智能M2模型，高性能通用模型", + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "inputPrice": 2, + "outputPrice": 20, + "defaultTemperature": 0.3, + "defaultTopP": 0.85, + "extra": {} + } + ] +} diff --git a/compute/providers/baidu.json b/compute/providers/baidu.json new file mode 100644 index 0000000..74b1039 --- /dev/null +++ b/compute/providers/baidu.json @@ -0,0 +1,158 @@ +{ + "id": "provider-baidu-001", + "provider": "baidu", + "label": "百度 文心", + "baseUrl": "https://qianfan.baidubce.com/v2", + "apiFormat": "openai-completions", + "apiKeyRef": "baidu", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "ernie-5.0-thinking-latest", + "displayName": "文心 ERNIE 5.0", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "百度最新旗舰模型，支持深度推理、多模态和工具调用，119K输入/64K输出", + "contextWindow": 128000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "vision", + "multimodal", + "tool_use" + ], + "inputPrice": 6, + "outputPrice": 24, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32768, + "inputPrice": 6, + "outputPrice": 24 + }, + { + "maxInputTokens": 131072, + "inputPrice": 10, + "outputPrice": 40 + } + ], + "thinkingMaxTokens": 60000 + } + }, + { + "modelName": "ernie-5.0", + "displayName": "文心 ERNIE 5.0 非思考", + "serviceType": [ + "chat" + ], + "description": "百度 ERNIE 5.0 旗舰模型，128K 上下文，119K 最大输入，最大输出 65,536 token", + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "vision", + "multimodal", + "tool_use" + ], + "inputPrice": 6, + "outputPrice": 24, + "defaultTemperature": 0.95, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32768, + "inputPrice": 6, + "outputPrice": 24 + }, + { + "maxInputTokens": 131072, + "inputPrice": 10, + "outputPrice": 40 + } + ] + } + }, + { + "modelName": "ernie-4.5-turbo-128k", + "displayName": "文心 ERNIE 4.5 Turbo", + "serviceType": [ + "chat" + ], + "description": "百度高性价比长上下文模型，128K 窗口，最大输出 12,288 token", + "contextWindow": 131072, + "maxOutputTokens": 12288, + "capabilities": [ + "chat", + "code", + "vision", + "long_context", + "fast" + ], + "inputPrice": 0.8, + "outputPrice": 3.2, + "defaultTemperature": 0.8, + "defaultTopP": 1, + "extra": { + "cacheHitPrice": 0.2 + } + }, + { + "modelName": "ernie-4.5-turbo-20260402", + "displayName": "文心 ERNIE 4.5 Turbo 20260402", + "serviceType": [ + "chat" + ], + "description": "百度 ERNIE 4.5 Turbo 20260402 快照，128K 上下文，最大输出 12,288 token", + "contextWindow": 131072, + "maxOutputTokens": 12288, + "capabilities": [ + "chat", + "code", + "vision", + "long_context", + "fast" + ], + "inputPrice": 0.8, + "outputPrice": 3.2, + "defaultTemperature": 0.8, + "defaultTopP": 1, + "extra": { + "cacheHitPrice": 0.2 + } + }, + { + "modelName": "ernie-x1.1", + "displayName": "文心 ERNIE X1.1", + "serviceType": [ + "reasoning" + ], + "description": "百度 ERNIE X1.1 深度思考模型，64K 上下文，最大输出 65,536 token", + "contextWindow": 65536, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "math", + "code", + "tool_use" + ], + "inputPrice": 1, + "outputPrice": 4, + "extra": { + "thinkingMaxTokens": 65536 + } + } + ] +} diff --git a/compute/providers/cohere.json b/compute/providers/cohere.json new file mode 100644 index 0000000..9d91f6e --- /dev/null +++ b/compute/providers/cohere.json @@ -0,0 +1,151 @@ +{ + "id": "provider-cohere-001", + "provider": "cohere", + "label": "Cohere", + "baseUrl": "https://api.cohere.ai/compatibility/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "cohere", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "embedding", + "rerank" + ], + "models": [ + { + "modelName": "embed-v5.0-fast", + "displayName": "Embed 5 Fast", + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Embed 5 Fast；当前 OpenAI 兼容接入仅声明文本向量能力", + "contextWindow": 128000, + "extra": { + "defaultDimension": 2048, + "pricingNotes": "Token 单价未由当前官方价格页公开确认；Model Vault 按实例计价。 Compatibility API 不接受 dimensions 参数；返回默认维度，多维度选择需原生 Embed API。" + } + }, + { + "modelName": "embed-v5.0-pro", + "displayName": "Embed 5 Pro", + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Embed 5 Pro；当前 OpenAI 兼容接入仅声明文本向量能力", + "contextWindow": 128000, + "extra": { + "defaultDimension": 2048, + "pricingNotes": "Token 单价未由当前官方价格页公开确认；Model Vault 按实例计价。 Compatibility API 不接受 dimensions 参数；返回默认维度，多维度选择需原生 Embed API。" + } + }, + { + "modelName": "command-a-plus-05-2026", + "displayName": "Command A+", + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "multilingual", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "Command A+", + "contextWindow": 128000, + "maxOutputTokens": 64000, + "extra": { + "pricingNotes": "官方文档声明限流额度内免费；生产 Model Vault 按实例单独计价，不预填 token 单价。", + "reasoning": { + "supportedEfforts": [ + "none", + "high" + ] + } + } + }, + { + "modelName": "command-a-03-2025", + "displayName": "Command A", + "serviceType": [ + "chat" + ], + "description": "Cohere Command A 文本模型，256K 上下文，支持 RAG 和工具调用", + "contextWindow": 256000, + "maxOutputTokens": 8000, + "capabilities": [ + "chat", + "code", + "tool_use", + "rag", + "long_context" + ], + "inputPrice": 2.5, + "outputPrice": 10, + "defaultTemperature": 0.3, + "defaultTopP": 0.75, + "extra": { + "pricingNotes": "Prices are per 1M tokens." + } + }, + { + "modelName": "command-r7b-12-2024", + "displayName": "Command R7B", + "serviceType": [ + "fast" + ], + "description": "Cohere 小型高速对话模型，适合高吞吐、低延迟场景", + "contextWindow": 128000, + "maxOutputTokens": 4000, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "rag", + "fast" + ], + "inputPrice": 0.0375, + "outputPrice": 0.15, + "defaultTemperature": 0.3, + "defaultTopP": 0.75, + "extra": { + "pricingNotes": "Prices are per 1M tokens." + } + }, + { + "modelName": "embed-v4.0", + "displayName": "Embed V4", + "serviceType": [ + "embedding" + ], + "description": "Cohere 最新 Embedding 模型，128K 上下文，多语言支持", + "contextWindow": 128000, + "maxOutputTokens": 0, + "capabilities": [ + "text_embedding", + "multilingual" + ], + "inputPrice": 0.12, + "outputPrice": 0, + "extra": { + "defaultDimension": 1536, + "pricingNotes": "Embedding models are priced by embedded tokens; price is per 1M tokens. Compatibility API 不接受 dimensions 参数；返回默认维度，多维度选择需原生 Embed API。" + } + } + ], + "tombstones": [ + "rerank-v3.5" + ] +} diff --git a/compute/providers/dashscope.json b/compute/providers/dashscope.json new file mode 100644 index 0000000..439a1b6 --- /dev/null +++ b/compute/providers/dashscope.json @@ -0,0 +1,943 @@ +{ + "id": "provider-dashscope-001", + "provider": "dashscope", + "label": "阿里 通义千问", + "baseUrl": "https://dashscope.aliyuncs.com/compatible-mode/v1", + "mediaBaseUrl": "https://dashscope.aliyuncs.com/api/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "dashscope", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "embedding", + "rerank", + "tts", + "asr", + "vision", + "image_gen", + "video_gen", + "voice_clone", + "omni", + "translation" + ], + "tombstones": [ + "qwen3.8-max-preview", + "qwen3.6-plus", + "qwen3.5-plus", + "qwen-max", + "qwen-plus", + "qwen-turbo", + "qwen3-max", + "qwen3-max-trans" + ], + "models": [ + { + "modelName": "qwen-image-3.0", + "displayName": "qwen-image-3.0", + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0", + "extra": { + "pricingNotes": "按图片计费，北京：输入 0.02 元/张；Pro 输出 1K 0.25 元、2K 0.5 元；普通版输出 0.18 元/张。" + } + }, + { + "modelName": "qwen-image-3.0-pro", + "displayName": "qwen-image-3.0-pro", + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0-pro", + "extra": { + "pricingNotes": "按图片计费，北京：输入 0.02 元/张；Pro 输出 1K 0.25 元、2K 0.5 元；普通版输出 0.18 元/张。" + } + }, + { + "modelName": "qwen3.7-text-rerank", + "displayName": "Qwen3.7 Text Rerank", + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Qwen3.7 Text Rerank" + }, + { + "modelName": "qwen3.7-text-embedding-flash", + "displayName": "qwen3.7-text-embedding-flash", + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "qwen3.7-text-embedding-flash", + "inputPrice": 0.125, + "outputPrice": 0 + }, + { + "modelName": "qwen3.7-text-embedding", + "displayName": "qwen3.7-text-embedding", + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "qwen3.7-text-embedding", + "inputPrice": 0.5, + "outputPrice": 0 + }, + { + "modelName": "qwen3.8-omni-flash", + "displayName": "Qwen3.8 Omni Flash", + "capabilities": [ + "chat", + "reasoning", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "Qwen3.8 Omni Flash", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "inputPrice": 0.8, + "outputPrice": 2.7, + "extra": { + "cachedInputPrice": 0.1 + } + }, + { + "modelName": "qwen3.8-max", + "displayName": "阿里云 Qwen3.8-Max", + "serviceType": [ + "chat", + "vision" + ], + "description": "通义千问3.8 Max 稳定版，支持视觉、视频理解和工具调用，100万上下文", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision", + "video_understanding" + ], + "inputPrice": 12, + "outputPrice": 36, + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "extra": { + "cachedInputPrice": 1.5, + "pricingNotes": "CNY per 1M tokens, Beijing region standard price.", + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + }, + "preserveThinkingDefault": true, + "thinkingMaxTokens": 262144 + } + }, + { + "modelName": "qwen3.8-flash", + "displayName": "阿里云 Qwen3.8-Flash", + "serviceType": [ + "chat", + "vision" + ], + "description": "通义千问3.8 Flash 稳定版，高速多模态推理，100万上下文", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision", + "video_understanding" + ], + "inputPrice": 0.8, + "outputPrice": 2.7, + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "extra": { + "cachedInputPrice": 0.1, + "pricingNotes": "CNY per 1M tokens, Beijing region standard price.", + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + }, + "preserveThinkingDefault": true, + "thinkingMaxTokens": 262144 + } + }, + { + "modelName": "qwen3.7-max", + "displayName": "阿里云 Qwen3.7-Max", + "serviceType": [ + "chat" + ], + "description": "通义千问3.7 Max，面向复杂推理与长程 Agent 任务的旗舰模型，100万上下文，支持内置工具和 Function Calling", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent" + ], + "inputPrice": 12, + "outputPrice": 36, + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 2.4 + } + }, + { + "modelName": "qwen3.7-plus", + "displayName": "阿里云 Qwen3.7-Plus", + "serviceType": [ + "chat", + "vision" + ], + "description": "通义千问3.7 Plus，多模态智能体模型，100万上下文，支持内置工具、Function Calling 和结构化输出", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "inputPrice": 2, + "outputPrice": 8, + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 256000, + "inputPrice": 2, + "outputPrice": 8, + "thinkingOutputPrice": 8 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 6, + "outputPrice": 24, + "thinkingOutputPrice": 24 + } + ], + "thinkingMaxTokens": 262144 + } + }, + { + "modelName": "qwen3.6-flash", + "displayName": "阿里云 Qwen3.6-Flash", + "serviceType": [ + "chat", + "vision" + ], + "description": "通义千问3.6 Flash，高性价比多模态模型，100万上下文，支持内置工具和 Function Calling", + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "fast" + ], + "inputPrice": 1.2, + "outputPrice": 7.2, + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 256000, + "inputPrice": 1.2, + "outputPrice": 7.2 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 4.8, + "outputPrice": 28.8 + } + ], + "thinkingMaxTokens": 131072 + } + }, + { + "modelName": "qwen-long", + "displayName": "阿里云 qwen-long", + "serviceType": [ + "chat" + ], + "description": "通义千问Long，支持1000万上下文，适合超长文档处理，降价97%", + "contextWindow": 10000000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "ultra_long_context", + "document_processing" + ], + "inputPrice": 0.5, + "outputPrice": 2, + "defaultTemperature": 1, + "defaultTopP": 0.8, + "extra": {} + }, + { + "modelName": "qwen3-vl-plus", + "displayName": "阿里云 Qwen3-VL-Plus", + "serviceType": [ + "vision" + ], + "description": "通义千问3代视觉模型Plus版，支持图像理解", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "ocr", + "chart_analysis", + "tool_use" + ], + "inputPrice": 1.5, + "outputPrice": 6, + "defaultTemperature": 0.7, + "defaultTopP": 0.8, + "extra": {} + }, + { + "modelName": "qwen3-vl-flash", + "displayName": "阿里云 Qwen3-VL-Flash", + "serviceType": [ + "vision" + ], + "description": "通义千问3代视觉模型Flash版，高性价比", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "fast", + "tool_use" + ], + "inputPrice": 0.8, + "outputPrice": 3, + "defaultTemperature": 0.7, + "defaultTopP": 0.8, + "extra": {} + }, + { + "modelName": "text-embedding-v3", + "displayName": "text-embedding-v3", + "serviceType": [ + "embedding" + ], + "description": "阿里云文本向量化", + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "chinese_optimized" + ], + "inputPrice": 0.7, + "extra": {} + }, + { + "modelName": "text-embedding-v4", + "displayName": "阿里云 text-embedding-v4", + "serviceType": [ + "embedding" + ], + "description": "阿里云百炼文本向量模型v4，支持多维度与多语种", + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag" + ], + "inputPrice": 0.5, + "extra": {} + }, + { + "modelName": "qwen3-rerank", + "displayName": "Qwen3-Rerank", + "serviceType": [ + "rerank" + ], + "description": "检索重排序，支持 100+ 语种", + "contextWindow": 120000, + "capabilities": [ + "rerank", + "semantic_reranking", + "multilingual" + ], + "inputPrice": 1, + "extra": {} + }, + { + "modelName": "cosyvoice-v2", + "displayName": "CosyVoice V2", + "serviceType": [ + "tts" + ], + "description": "通义语音合成，支持中英文", + "capabilities": [ + "tts", + "multilingual", + "chinese_optimized" + ], + "extra": { + "voices": [ + "longxiaochun", + "longxiaoxia", + "longlaotie", + "longshu" + ], + "supportedAudioFormats": [ + "mp3", + "wav", + "pcm" + ] + } + }, + { + "modelName": "paraformer-v2", + "displayName": "Paraformer V2", + "serviceType": [ + "asr" + ], + "description": "通义语音识别", + "capabilities": [ + "asr", + "chinese_optimized" + ], + "extra": { + "maxAudioLength": 3600, + "supportedInputFormats": [ + "wav", + "mp3", + "pcm", + "ogg" + ], + "streamingSupported": true + } + }, + { + "modelName": "wan2.7-image-pro", + "displayName": "通义万相 2.7 Pro", + "serviceType": [ + "image_gen" + ], + "description": "万相最新旗舰文生图模型，支持 4K 分辨率、thinking_mode 智能思考、色彩调色板控制", + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "768*768", + "1024*1024", + "1280*1280", + "1024*1792", + "1792*1024", + "2048*2048", + "4096*4096" + ], + "sizeShortcuts": { + "1K": "1024*1024", + "2K": "2048*2048", + "4K": "4096*4096" + }, + "maxN": 4, + "supportsThinkingMode": true, + "supportsColorPalette": true + } + }, + { + "modelName": "wan2.7-image", + "displayName": "通义万相 2.7", + "serviceType": [ + "image_gen" + ], + "description": "万相最新标准文生图模型，高画质，支持 thinking_mode", + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "768*768", + "1024*1024", + "1280*1280", + "1024*1792", + "1792*1024" + ], + "maxN": 4, + "supportsThinkingMode": true + } + }, + { + "modelName": "wan2.6-t2i", + "displayName": "通义万相 2.6 文生图", + "serviceType": [ + "image_gen" + ], + "description": "万相 2.6 文生图专用模型，稳定可靠，支持 prompt_extend 智能改写", + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "1024*1024", + "1280*1280", + "720*1280", + "1280*720", + "768*1152", + "1152*768" + ], + "maxN": 4, + "supportsNegativePrompt": true, + "supportsPromptExtend": true + } + }, + { + "modelName": "wan2.2-t2i-plus", + "displayName": "通义万相 2.2 Plus", + "serviceType": [ + "image_gen" + ], + "description": "万相 2.2 增强版，画质与速度平衡", + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "1024*1024", + "720*1280", + "1280*720" + ], + "maxN": 4, + "supportsNegativePrompt": true + } + }, + { + "modelName": "wan2.2-t2i-flash", + "displayName": "通义万相 2.2 Flash", + "serviceType": [ + "image_gen" + ], + "description": "万相 2.2 快速版，生成速度极快，适合批量和实时场景", + "capabilities": [ + "image_generation", + "chinese_optimized", + "fast" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "1024*1024", + "720*1280", + "1280*720" + ], + "maxN": 4, + "supportsNegativePrompt": true + } + }, + { + "modelName": "qwen-image-2.0-pro", + "displayName": "千问图像 2.0 Pro", + "serviceType": [ + "image_gen" + ], + "description": "千问图像生成 2.0 Pro 旗舰版，高画质多风格", + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "512*512", + "1024*1024", + "1280*1280", + "1664*928", + "928*1664" + ], + "maxN": 4 + } + }, + { + "modelName": "qwen-image-2.0", + "displayName": "千问图像 2.0", + "serviceType": [ + "image_gen" + ], + "description": "千问图像生成 2.0 标准版", + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "extra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "512*512", + "1024*1024", + "1280*1280", + "2048*2048" + ], + "maxN": 4 + } + }, + { + "modelName": "wan3.0-video-prime", + "displayName": "通义万相 3.0 视频生成高速版", + "serviceType": [ + "video_gen" + ], + "description": "万相 3.0 全能参考视频生成高速版，能力对齐标准版并显著提升端到端生成速度", + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "chinese_optimized", + "high_quality", + "fast" + ], + "extra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 30, + "supportsIntelligentDuration": true, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ], + "supportedRatios": [ + "adaptive", + "16:9", + "4:3", + "1:1", + "3:4", + "9:16" + ], + "mediaInputTypes": [ + "first_frame", + "last_frame", + "reference_image", + "reference_video", + "reference_audio", + "file", + "link" + ], + "maxReferenceImages": 10, + "maxReferenceVideos": 5, + "maxReferenceAudios": 5, + "supportsOutputAudio": true, + "supportsSeed": true, + "supportsPromptExtend": true, + "supportsWatermark": true + } + }, + { + "modelName": "wan3.0-video", + "displayName": "通义万相 3.0 视频生成", + "serviceType": [ + "video_gen" + ], + "description": "万相 3.0 全能参考视频生成标准版，统一支持文生、首帧/首尾帧图生及多类型参考生视频", + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "chinese_optimized", + "high_quality" + ], + "extra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 30, + "supportsIntelligentDuration": true, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ], + "supportedRatios": [ + "adaptive", + "16:9", + "4:3", + "1:1", + "3:4", + "9:16" + ], + "mediaInputTypes": [ + "first_frame", + "last_frame", + "reference_image", + "reference_video", + "reference_audio", + "file", + "link" + ], + "maxReferenceImages": 10, + "maxReferenceVideos": 5, + "maxReferenceAudios": 5, + "supportsOutputAudio": true, + "supportsSeed": true, + "supportsPromptExtend": true, + "supportsWatermark": true + } + }, + { + "modelName": "wan2.6-t2v", + "displayName": "通义万相 2.6 文生视频", + "serviceType": [ + "video_gen" + ], + "description": "万相 2.6 文生视频模型，支持文生视频和图生视频", + "capabilities": [ + "video_generation", + "chinese_optimized" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "720p", + "1080p" + ], + "endpoint": "/services/aigc/video-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks" + } + }, + { + "modelName": "cosyvoice-clone", + "displayName": "CosyVoice 声音复刻", + "serviceType": [ + "voice_clone" + ], + "description": "通义声音复刻，少量样本即可克隆", + "capabilities": [ + "voice_clone", + "chinese_optimized", + "few_shot" + ], + "extra": { + "maxCloneSamples": 5, + "minCloneDuration": 3 + } + }, + { + "modelName": "qwen-omni-turbo", + "displayName": "Qwen-Omni-Turbo", + "serviceType": [ + "omni" + ], + "description": "通义全能模型，支持音视频文本多模态", + "contextWindow": 32768, + "maxOutputTokens": 2048, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "multimodal", + "tool_use" + ], + "defaultTemperature": 0.7, + "defaultTopP": 0.01, + "extra": {} + }, + { + "modelName": "qwen-mt-plus", + "displayName": "Qwen-MT-Plus", + "serviceType": [ + "translation" + ], + "description": "通义千问 Qwen-MT 旗舰翻译模型，支持 92 种语言互译，适合专业领域、正式文书和技术报告", + "contextWindow": 16384, + "maxOutputTokens": 8192, + "capabilities": [ + "translation", + "multilingual", + "chinese_optimized", + "terminology" + ], + "inputPrice": 1.8, + "outputPrice": 5.4, + "extra": { + "supportedLanguageCount": 92, + "supportsIncrementalStreaming": false + } + }, + { + "modelName": "happyhorse-1.0-t2v", + "displayName": "快乐马 1.0 文生视频", + "serviceType": [ + "video_gen" + ], + "description": "阿里 HappyHorse 文生视频，原生音视频联合生成，最长 15s，支持多语种口型", + "capabilities": [ + "video_generation", + "text_to_video", + "chinese_optimized", + "high_quality" + ], + "extra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 15, + "supportedResolutions": [ + "720P", + "1080P" + ] + } + }, + { + "modelName": "happyhorse-1.0-i2v", + "displayName": "快乐马 1.0 图生视频", + "serviceType": [ + "video_gen" + ], + "description": "阿里 HappyHorse 图生视频（首帧），宽高比自动跟随输入首帧", + "capabilities": [ + "video_generation", + "image_to_video", + "chinese_optimized" + ], + "extra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 15, + "supportedResolutions": [ + "720P", + "1080P" + ] + } + }, + { + "modelName": "happyhorse-1.0-r2v", + "displayName": "快乐马 1.0 参考生视频", + "serviceType": [ + "video_gen" + ], + "description": "阿里 HappyHorse 参考生视频，支持最多 9 张参考图（多主体/元素一致性）", + "capabilities": [ + "video_generation", + "reference_to_video", + "multi_image", + "chinese_optimized" + ], + "extra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxReferenceImages": 9, + "maxVideoDuration": 15, + "supportedResolutions": [ + "720P", + "1080P" + ] + } + } + ] +} diff --git a/compute/providers/deepseek.json b/compute/providers/deepseek.json new file mode 100644 index 0000000..eb69bda --- /dev/null +++ b/compute/providers/deepseek.json @@ -0,0 +1,219 @@ +{ + "id": "provider-deepseek-001", + "provider": "deepseek", + "label": "DeepSeek", + "baseUrl": "https://api.deepseek.com", + "apiFormat": "openai-completions", + "apiKeyRef": "deepseek", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning", + "vision" + ], + "tombstones": [ + "deepseek-chat", + "deepseek-reasoner" + ], + "models": [ + { + "modelName": "deepseek-flash", + "displayName": "DeepSeek V4.1 Flash", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "DeepSeek V4.1 Flash，支持文本与图像输入，1M 上下文，最大输出 384K，并发上限 2500", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "long_context", + "multilingual", + "tool_use" + ], + "inputPrice": 1, + "outputPrice": 4, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cacheHitPrice": 0.02, + "concurrencyLimit": 2500, + "supportsThinking": true, + "thinkingDefault": true, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + }, + "pricingTiers": [ + { + "condition": "idle", + "inputPrice": 1, + "outputPrice": 4, + "cacheHitPrice": 0.02 + }, + { + "condition": "peak", + "inputPrice": 2, + "outputPrice": 8, + "cacheHitPrice": 0.04 + } + ], + "pricingNotes": "空闲时段为北京时间工作日 09:00-12:00、14:00-18:00 之外的时段；高峰时段价格为输入 2 元、输出 8 元、缓存命中输入 0.04 元/百万 tokens。", + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": "non-thinking-only" + } + } + }, + { + "modelName": "deepseek-v4-flash-vision-exp", + "displayName": "DeepSeek V4 Flash Vision Exp", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "DeepSeek V4 Flash 实验性视觉版本，支持文本与图像输入，文本能力与 V4 Flash 一致，1M 上下文，并发上限 2500", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "long_context", + "multilingual", + "tool_use" + ], + "inputPrice": 1.008, + "outputPrice": 2.016, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cacheHitPrice": 0.02016, + "concurrencyLimit": 2500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": false + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + } + }, + { + "modelName": "deepseek-v4-flash", + "displayName": "DeepSeek V4 Flash", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "DeepSeek V4 Flash，284B 总参数 / 13B 激活 MoE，1M 上下文，默认思考模式，支持非思考模式，并发上限 2500", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "long_context", + "multilingual", + "tool_use" + ], + "inputPrice": 1.008, + "outputPrice": 2.016, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cacheHitPrice": 0.02016, + "concurrencyLimit": 2500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": "non-thinking-only" + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + } + }, + { + "modelName": "deepseek-v4-pro", + "displayName": "DeepSeek V4 Pro", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "DeepSeek V4 Pro，1.6T 总参数 / 49B 激活 MoE，1M 上下文，默认思考模式，支持非思考模式，并发上限 500", + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "long_context", + "multilingual", + "tool_use", + "math" + ], + "inputPrice": 3.132, + "outputPrice": 6.264, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cacheHitPrice": 0.0261, + "concurrencyLimit": 500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": "non-thinking-only" + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + } + } + ] +} diff --git a/compute/providers/github-copilot.json b/compute/providers/github-copilot.json new file mode 100644 index 0000000..4327880 --- /dev/null +++ b/compute/providers/github-copilot.json @@ -0,0 +1,28 @@ +{ + "id": "provider-github-copilot-plan-001", + "provider": "github-copilot", + "brandGroup": "github", + "label": "GitHub Copilot 订阅", + "baseUrl": "https://api.githubcopilot.com", + "apiFormat": "openai-completions", + "apiKeyRef": "github-copilot-session-token", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "accessMode": "coding-plan", + "credentialSource": "github-copilot", + "requiredClientVersion": "10.0.132", + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://github.com/settings/copilot/features" + } + }, + "services": [ + "chat", + "reasoning" + ], + "models": [] +} diff --git a/compute/providers/google.json b/compute/providers/google.json new file mode 100644 index 0000000..683b3b2 --- /dev/null +++ b/compute/providers/google.json @@ -0,0 +1,323 @@ +{ + "id": "provider-google-001", + "provider": "google", + "label": "Google Gemini", + "baseUrl": "https://generativelanguage.googleapis.com/v1beta", + "apiFormat": "google-generative-ai", + "apiKeyRef": "google", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "vision", + "embedding" + ], + "models": [ + { + "modelName": "gemini-embedding-2", + "displayName": "Gemini Embedding 2", + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "vision", + "audio_understanding", + "video_understanding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Gemini Embedding 2，原生多模态嵌入，8192 token 输入，共享文本/图像/音视频向量空间", + "extra": { + "dimensions": [ + 768, + 1536, + 3072 + ], + "defaultDimension": 3072 + } + }, + { + "modelName": "gemini-3.8-flash", + "displayName": "Gemini 3.8 Flash", + "serviceType": [ + "chat" + ], + "description": "Google 最新稳定版 Flash，面向长程编程和智能体任务，1M 上下文", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.75, + "outputPrice": 3.75, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cachePricing": { + "inputCacheRead": 0.075 + }, + "reasoning": { + "supportedEfforts": ["low", "medium", "high"], + "defaultEffort": "medium" + }, + "pricingNotes": "Standard paid tier introductory price through 2026-12-31; standard rates rise on 2027-01-01." + } + }, + { + "modelName": "gemini-3.7-flash", + "displayName": "Gemini 3.7 Flash", + "serviceType": [ + "chat" + ], + "description": "Google 稳定版 Flash，面向日常编程和多步工具调用，1M 上下文", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.75, + "outputPrice": 3.75, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cachePricing": { + "inputCacheRead": 0.075 + }, + "reasoning": { + "supportedEfforts": ["low", "medium", "high"], + "defaultEffort": "medium" + }, + "pricingNotes": "Standard paid tier introductory price through 2026-12-31; standard rates rise on 2027-01-01." + } + }, + { + "modelName": "gemini-3.5-flash-lite", + "displayName": "Gemini 3.5 Flash-Lite", + "serviceType": [ + "chat", + "fast" + ], + "description": "Google 低延迟轻量模型，支持多模态输入，1M 上下文", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.3, + "outputPrice": 2.5, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cachePricing": { + "inputCacheRead": 0.03 + } + } + }, + { + "modelName": "gemini-3.5-flash", + "displayName": "Gemini 3.5 Flash", + "serviceType": [ + "chat" + ], + "description": "Google Gemini 3.5 Flash 稳定版，高速多模态模型，1M 上下文", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 1.5, + "outputPrice": 9, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cachePricing": { + "inputCacheRead": 0.75 + }, + "pricingNotes": "Standard paid tier text/image/video price; audio input uses a separate Google pricing row." + } + }, + { + "modelName": "gemini-3.1-flash-lite", + "displayName": "Gemini 3.1 Flash-Lite", + "serviceType": [ + "fast" + ], + "description": "Google Gemini 3.1 Flash-Lite 稳定版，低成本高吞吐多模态模型，替代已关闭的 Preview 版本", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.25, + "outputPrice": 1.5, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "audioInputPrice": 0.5, + "cachePricing": { + "inputCacheRead": 0.125, + "audioInputCacheRead": 0.25 + } + } + }, + { + "modelName": "gemini-3.1-pro-preview", + "displayName": "Gemini 3.1 Pro Preview", + "serviceType": [ + "chat" + ], + "description": "Google Gemini 3.1 Pro Preview，面向复杂推理、编码和智能体工作流", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "inputPrice": 2, + "outputPrice": 12, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "pricingNotes": "Standard paid tier text/image/video price for prompts <= 200K tokens; prompts > 200K are $4 input and $18 output per 1M tokens.", + "cachePricing": { + "inputCacheRead": 0.2, + "inputCacheReadOver200k": 0.4, + "storagePerMillionTokensPerHour": 4.5 + } + } + }, + { + "modelName": "gemini-3-flash-preview", + "displayName": "Gemini 3 Flash Preview", + "serviceType": [ + "chat" + ], + "description": "Google Gemini 3 Flash Preview，高速多模态推理模型", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.5, + "outputPrice": 3, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "pricingNotes": "Standard paid tier text/image/video price; audio input is $1.00 per 1M tokens.", + "cachePricing": { + "inputCacheRead": 0.05, + "audioInputCacheRead": 0.1, + "storagePerMillionTokensPerHour": 1 + } + } + }, + { + "modelName": "gemini-2.5-pro", + "displayName": "Gemini 2.5 Pro", + "serviceType": [ + "chat" + ], + "description": "Google Gemini 2.5 Pro，1,048,576 token 上下文，多模态", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "inputPrice": 1.25, + "outputPrice": 10, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": {} + }, + { + "modelName": "gemini-2.5-flash", + "displayName": "Gemini 2.5 Flash", + "serviceType": [ + "chat" + ], + "description": "Google 高性价比模型，1,048,576 token 上下文，速度快", + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.3, + "outputPrice": 2.5, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": {} + }, + { + "modelName": "text-embedding-005", + "displayName": "Text Embedding 005", + "serviceType": [ + "embedding" + ], + "description": "Google 文本嵌入模型，768维度", + "contextWindow": 2048, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "inputPrice": 0.1, + "extra": {} + } + ], + "tombstones": [ + "gemini-3.1-flash-lite-preview" + ] +} diff --git a/compute/providers/kling.json b/compute/providers/kling.json new file mode 100644 index 0000000..13f2d37 --- /dev/null +++ b/compute/providers/kling.json @@ -0,0 +1,104 @@ +{ + "id": "provider-kling-001", + "provider": "kling", + "label": "快手 可灵", + "baseUrl": "https://api.klingai.com/v1", + "apiFormat": "kling-task-api", + "apiKeyRef": "kling", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "video_gen" + ], + "models": [ + { + "modelName": "kling-v3", + "displayName": "可灵 3.0", + "serviceType": [ + "video_gen" + ], + "capabilities": [ + "video_generation", + "image_to_video" + ], + "description": "可灵 3.0，官方 legacy /v1/videos/text2video 协议已支持该 model_name" + }, + { + "modelName": "kling-v2-5-turbo", + "displayName": "可灵 V2.5 Turbo", + "serviceType": ["video_gen"], + "description": "可灵最新视频生成模型，高性价比", + "capabilities": [ + "video_generation", + "chinese_optimized", + "high_quality", + "fast" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "720p", + "1080p" + ], + "pricePerGeneration": 1.5 + } + }, + { + "modelName": "kling-v2-5-turbo-pro", + "displayName": "可灵 V2.5 Turbo Pro", + "serviceType": ["video_gen"], + "description": "可灵高品质视频生成", + "capabilities": [ + "video_generation", + "chinese_optimized", + "ultra_quality" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "1080p", + "4k" + ], + "pricePerGeneration": 2.5 + } + }, + { + "modelName": "kling-v2", + "displayName": "可灵 V2", + "serviceType": ["video_gen"], + "description": "可灵文生视频旗舰模型", + "capabilities": [ + "video_generation", + "chinese_optimized", + "high_quality" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "720p", + "1080p" + ] + } + }, + { + "modelName": "kling-v2-master", + "displayName": "可灵 V2 Master", + "serviceType": ["video_gen"], + "description": "可灵最高品质视频生成", + "capabilities": [ + "video_generation", + "chinese_optimized", + "ultra_quality" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "1080p", + "4k" + ] + } + } + ] +} diff --git a/compute/providers/local-whisper.json b/compute/providers/local-whisper.json new file mode 100644 index 0000000..f8c58ad --- /dev/null +++ b/compute/providers/local-whisper.json @@ -0,0 +1,38 @@ +{ + "id": "provider-local-whisper", + "provider": "local-whisper", + "label": "本地 Whisper", + "baseUrl": "http://localhost:8080", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "asr" + ], + "models": [ + { + "modelName": "whisper-large-v3", + "displayName": "Whisper Large V3", + "serviceType": ["asr"], + "description": "本地部署的 Whisper 大模型", + "capabilities": [ + "asr", + "multilingual" + ], + "extra": { + "maxAudioLength": 7200, + "supportedInputFormats": [ + "wav", + "mp3", + "flac", + "ogg", + "webm" + ], + "streamingSupported": true + } + } + ] +} diff --git a/compute/providers/minimax.json b/compute/providers/minimax.json new file mode 100644 index 0000000..51c7b81 --- /dev/null +++ b/compute/providers/minimax.json @@ -0,0 +1,728 @@ +{ + "id": "provider-minimax-001", + "provider": "minimax", + "label": "MiniMax", + "baseUrl": "https://api.minimaxi.com/anthropic", + "apiFormat": "anthropic-messages", + "mediaBaseUrl": "https://api.minimaxi.com/v1", + "apiKeyRef": "minimax", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "fast", + "reasoning", + "vision", + "image_gen", + "tts", + "video_gen", + "music_gen" + ], + "models": [ + { + "modelName": "MiniMax-M3", + "displayName": "MiniMax M3", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "MiniMax M3，前沿多模态 Coding 与 Agent 模型，1M 上下文", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "vision", + "long_context" + ], + "inputPrice": 2.1, + "outputPrice": 8.4, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 512000, + "inputPrice": 2.1, + "outputPrice": 8.4, + "cacheReadPrice": 0.42 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 4.2, + "outputPrice": 16.8, + "cacheReadPrice": 0.84 + } + ], + "priorityPricingTiers": [ + { + "maxInputTokens": 512000, + "inputPrice": 3.15, + "outputPrice": 12.6, + "cacheReadPrice": 0.63 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 6.3, + "outputPrice": 25.2, + "cacheReadPrice": 1.26 + } + ], + "pricingNotes": "MiniMax 国内开放平台按量计费永久五折价格。" + } + }, + { + "modelName": "MiniMax-M2.7", + "displayName": "MiniMax M2.7", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "MiniMax 新一代旗舰模型，擅长复杂 Agent、软件工程与专业办公任务", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "inputPrice": 2.1, + "outputPrice": 8.4, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.42, + "cacheWritePrice": 2.625, + "outputSpeedTps": 60 + } + }, + { + "modelName": "MiniMax-M2.7-highspeed", + "displayName": "MiniMax M2.7 高速版", + "serviceType": [ + "fast" + ], + "description": "MiniMax M2.7 极速版，效果一致，输出更快，适合低延迟场景", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "inputPrice": 4.2, + "outputPrice": 16.8, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.42, + "cacheWritePrice": 2.625, + "outputSpeedTps": 100 + } + }, + { + "modelName": "MiniMax-M2.5", + "displayName": "MiniMax M2.5", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "MiniMax M2.5，204.8K 上下文，支持深度推理和工具调用", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "long_context" + ], + "inputPrice": 2.1, + "outputPrice": 8.4, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.21, + "cacheWritePrice": 2.625, + "outputSpeedTps": 60 + } + }, + { + "modelName": "MiniMax-M2.5-highspeed", + "displayName": "MiniMax M2.5 高速版", + "serviceType": [ + "fast" + ], + "description": "MiniMax M2.5 高速推理版本(100TPS)，适合低延迟场景", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "inputPrice": 4.2, + "outputPrice": 16.8, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.21, + "cacheWritePrice": 2.625, + "outputSpeedTps": 100 + } + }, + { + "modelName": "MiniMax-M2.1", + "displayName": "MiniMax M2.1", + "serviceType": [ + "chat" + ], + "description": "MiniMax 上一代主力模型，200K 上下文，性价比优秀", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "code", + "tool_use", + "reasoning" + ], + "inputPrice": 2.1, + "outputPrice": 8.4, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.21, + "cacheWritePrice": 2.625, + "outputSpeedTps": 60 + } + }, + { + "modelName": "MiniMax-M2.1-highspeed", + "displayName": "MiniMax M2.1 高速版", + "serviceType": [ + "fast" + ], + "description": "MiniMax M2.1 高速推理版本，适合低延迟场景", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "inputPrice": 4.2, + "outputPrice": 16.8, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.21, + "cacheWritePrice": 2.625, + "outputSpeedTps": 100 + } + }, + { + "modelName": "MiniMax-M2", + "displayName": "MiniMax M2", + "serviceType": [ + "chat" + ], + "description": "MiniMax M2，优化编码和 Agent 工作流", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "inputPrice": 2.1, + "outputPrice": 8.4, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "legacy": true, + "cacheReadPrice": 0.21, + "cacheWritePrice": 2.625 + } + }, + { + "modelName": "M2-her", + "displayName": "MiniMax M2 Her", + "serviceType": [ + "chat" + ], + "description": "MiniMax 角色扮演对话模型，专为多轮角色对话设计", + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "roleplay" + ], + "inputPrice": 2.1, + "outputPrice": 8.4, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheReadPrice": 0.21, + "cacheWritePrice": 2.625 + } + }, + { + "modelName": "image-01", + "displayName": "MiniMax Image 01", + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图/图生图模型，支持主题参考（角色一致性），最长 1500 字符 prompt", + "capabilities": [ + "image_generation", + "subject_reference", + "chinese_optimized" + ], + "extra": { + "supportedAspectRatios": [ + "1:1", + "16:9", + "9:16", + "4:3", + "3:4", + "3:2", + "2:3", + "21:9" + ], + "maxPromptLength": 1500, + "maxBatchSize": 9, + "responseFormats": [ + "url", + "base64" + ], + "urlExpiry": "24h", + "pricePerImage": 0.025 + } + }, + { + "modelName": "image-01-live", + "displayName": "MiniMax Image 01 Live", + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图模型（Live 版），支持画风控制（漫画/元气/中世纪/水彩）", + "capabilities": [ + "image_generation", + "style_control", + "subject_reference", + "chinese_optimized" + ], + "extra": { + "supportedAspectRatios": [ + "1:1", + "16:9", + "9:16", + "4:3", + "3:4", + "3:2", + "2:3" + ], + "supportedStyles": [ + "漫画", + "元气", + "中世纪", + "水彩" + ], + "maxPromptLength": 1500, + "maxBatchSize": 9, + "responseFormats": [ + "url", + "base64" + ], + "urlExpiry": "24h" + } + }, + { + "modelName": "speech-2.8-hd", + "displayName": "MiniMax Speech 2.8 HD", + "serviceType": [ + "tts" + ], + "description": "MiniMax 最新高清语音合成模型，支持 40+ 语言、100+ 预置音色，毫秒级延迟", + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone", + "chinese_optimized" + ], + "extra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ], + "pricePerMillionCharacters": 350 + } + }, + { + "modelName": "speech-2.8-turbo", + "displayName": "MiniMax Speech 2.8 Turbo", + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成模型，速度优先，适合实时场景", + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast", + "chinese_optimized" + ], + "extra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ], + "pricePerMillionCharacters": 200 + } + }, + { + "modelName": "speech-2.6-hd", + "displayName": "MiniMax Speech 2.6 HD", + "serviceType": [ + "tts" + ], + "description": "MiniMax 高清语音合成模型（2.6 版），音质优秀，生成速度快", + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone", + "chinese_optimized" + ], + "extra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ], + "pricePerMillionCharacters": 350 + } + }, + { + "modelName": "speech-2.6-turbo", + "displayName": "MiniMax Speech 2.6 Turbo", + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成模型（2.6 版），低延迟响应", + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast", + "chinese_optimized" + ], + "extra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ], + "pricePerMillionCharacters": 200 + } + }, + { + "modelName": "speech-02-hd", + "displayName": "MiniMax Speech 02 HD", + "serviceType": [ + "tts" + ], + "description": "MiniMax 高清语音合成模型（历史版本），韵律稳定性优秀", + "capabilities": [ + "tts", + "multilingual", + "voice_clone", + "chinese_optimized" + ], + "extra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ], + "legacy": true, + "pricePerMillionCharacters": 350 + } + }, + { + "modelName": "speech-02-turbo", + "displayName": "MiniMax Speech 02 Turbo", + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成模型（历史版本），增强小语种支持", + "capabilities": [ + "tts", + "multilingual", + "fast", + "chinese_optimized" + ], + "extra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ], + "legacy": true, + "pricePerMillionCharacters": 200 + } + }, + { + "modelName": "MiniMax-Hailuo-2.3", + "displayName": "海螺视频 2.3", + "serviceType": [ + "video_gen" + ], + "description": "MiniMax 最新视频生成模型，支持文生视频和图生视频，支持 15 种运镜指令控制", + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "camera_control", + "chinese_optimized", + "high_quality" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "768p", + "1080p" + ], + "maxPromptLength": 2000, + "asyncMode": true + } + }, + { + "modelName": "MiniMax-Hailuo-2.3-Fast", + "displayName": "海螺视频 2.3 快速版", + "serviceType": [ + "video_gen" + ], + "description": "海螺视频快速版本（仅图生视频），生成速度更快", + "capabilities": [ + "video_generation", + "image_to_video", + "camera_control", + "fast", + "chinese_optimized" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "768p", + "1080p" + ], + "asyncMode": true + } + }, + { + "modelName": "T2V-01-Director", + "displayName": "海螺视频 Director", + "serviceType": [ + "video_gen" + ], + "description": "导演级文生视频模型，精准运镜控制（历史版本）", + "capabilities": [ + "video_generation", + "text_to_video", + "camera_control", + "chinese_optimized" + ], + "extra": { + "maxVideoDuration": 6, + "supportedResolutions": [ + "720p" + ], + "asyncMode": true, + "legacy": true + } + }, + { + "modelName": "MiniMax-Hailuo-02", + "displayName": "海螺视频 02", + "serviceType": [ + "video_gen" + ], + "description": "海螺视频 02，原生 1080P 生成，指令跟随能力强（历史版本）", + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "chinese_optimized" + ], + "extra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "512p", + "768p", + "1080p" + ], + "asyncMode": true, + "legacy": true + } + }, + { + "modelName": "S2V-01", + "displayName": "海螺视频 主体参考", + "serviceType": [ + "video_gen" + ], + "description": "主体参考视频生成模型，基于参考图保持角色一致性", + "capabilities": [ + "video_generation", + "subject_reference", + "chinese_optimized" + ], + "extra": { + "asyncMode": true + } + }, + { + "modelName": "music-2.6", + "displayName": "MiniMax Music 2.6", + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成模型，支持音乐灵感描述和歌词输入，可生成带人声或纯音乐；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务", + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "extra": { + "maxPromptLength": 2000, + "maxLyricsLength": 3500, + "outputFormats": [ + "hex", + "url" + ], + "supportedLyricsStructures": [ + "[Intro]", + "[Verse]", + "[Pre Chorus]", + "[Chorus]", + "[Interlude]", + "[Bridge]", + "[Outro]", + "[Hook]", + "[Solo]", + "[Inst]" + ], + "pricePerSongUpToFiveMinutes": 1 + } + }, + { + "modelName": "music-2.5+", + "displayName": "MiniMax Music 2.5+", + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成模型（2.5+ 版），音质提升；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务", + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "extra": { + "outputFormats": [ + "hex", + "url" + ], + "pricePerSongUpToFiveMinutes": 1 + } + }, + { + "modelName": "music-2.5", + "displayName": "MiniMax Music 2.5", + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成模型（2.5 版）；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务", + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "extra": { + "outputFormats": [ + "hex", + "url" + ], + "pricePerSongUpToFiveMinutes": 1 + } + }, + { + "modelName": "music-cover", + "displayName": "MiniMax Music Cover", + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 翻唱模型，基于参考音频生成翻唱版本，支持修改歌词；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务", + "capabilities": [ + "music_generation", + "cover", + "chinese_optimized" + ], + "extra": { + "maxAudioDuration": 360, + "maxAudioSize": "50MB", + "supportedFormats": [ + "mp3", + "wav", + "flac" + ] + } + } + ], + "tombstones": [ + "MiniMax-Text-01" + ] +} diff --git a/compute/providers/mistral.json b/compute/providers/mistral.json new file mode 100644 index 0000000..e88a646 --- /dev/null +++ b/compute/providers/mistral.json @@ -0,0 +1,125 @@ +{ + "id": "provider-mistral-001", + "provider": "mistral", + "label": "Mistral AI", + "baseUrl": "https://api.mistral.ai/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "mistral", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "fast" + ], + "models": [ + { + "modelName": "mistral-medium-3-5", + "displayName": "Mistral Medium 3.5", + "contextWindow": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "Mistral Medium 3.5，官方 API ID mistral-medium-3-5，256K 上下文", + "inputPrice": 1.5, + "outputPrice": 7.5 + }, + { + "modelName": "mistral-medium-latest", + "displayName": "Mistral Medium 3.5", + "serviceType": [ + "chat" + ], + "description": "Mistral 前沿多模态模型，面向 Agentic 和 Coding 场景优化，256K 上下文", + "contextWindow": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 1.5, + "outputPrice": 7.5, + "defaultTemperature": 0.7, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "mistral-large-latest", + "displayName": "Mistral Large 3", + "serviceType": [ + "chat" + ], + "description": "Mistral 旗舰模型，256K 上下文，支持视觉和工具调用", + "contextWindow": 256000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "inputPrice": 0.5, + "outputPrice": 1.5, + "defaultTemperature": 0.7, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "mistral-small-latest", + "displayName": "Mistral Small 4", + "serviceType": [ + "fast" + ], + "description": "Mistral Small 4，指令、推理与编码混合模型，256K 上下文", + "contextWindow": 262144, + "capabilities": [ + "chat", + "code", + "vision", + "reasoning", + "fast", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 0.15, + "outputPrice": 0.6, + "defaultTemperature": 0.7, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "codestral-latest", + "displayName": "Codestral", + "serviceType": [ + "chat" + ], + "description": "Mistral 专业代码模型，支持 Fill-in-the-Middle 补全", + "contextWindow": 128000, + "maxOutputTokens": 32768, + "capabilities": [ + "code", + "fill_in_the_middle" + ], + "inputPrice": 0.3, + "outputPrice": 0.9, + "defaultTemperature": 0.7, + "defaultTopP": 1, + "extra": {} + } + ] +} diff --git a/compute/providers/moonshot.json b/compute/providers/moonshot.json new file mode 100644 index 0000000..fdc956c --- /dev/null +++ b/compute/providers/moonshot.json @@ -0,0 +1,419 @@ +{ + "id": "provider-moonshot-001", + "provider": "moonshot", + "label": "月之暗面 Kimi", + "baseUrl": "https://api.moonshot.cn/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "moonshot", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning", + "vision" + ], + "tombstones": [ + "kimi-k2", + "kimi-k2-0905-preview", + "kimi-k2-0711-preview", + "kimi-k2-turbo-preview", + "kimi-k2-thinking", + "kimi-k2-thinking-turbo", + "kimi-latest", + "kimi-thinking-preview" + ], + "models": [ + { + "modelName": "kimi-k3", + "displayName": "Kimi K3", + "contextWindow": 1048576, + "maxOutputTokens": 1048576, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding" + ], + "serviceType": [ + "reasoning" + ], + "description": "Kimi K3，1M 上下文，始终开启思考，low/high/max 三档（默认 max）；输出上限与输入共享窗口，默认 max_completion_tokens=131072", + "extra": { + "thinkingOnly": true, + "thinking": { + "default": "enabled", + "disableSupported": false, + "preserveReasoningContent": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + }, + "pricingNotes": "CNY per 1M tokens: input 20, output 100, cache hit 2. Cache writes: TTL 5min 20, TTL 1h 40; recorded separately from cache-hit billing.", + "samplingParametersDeprecated": true, + "cacheHitPrice": 2 + }, + "inputPrice": 20, + "outputPrice": 100 + }, + { + "modelName": "kimi-k2.7-code", + "displayName": "Kimi K2.7 Code", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "月之暗面 Kimi K2.7 Code，当前最强 Coding 模型，256K 上下文，支持文本/图片/视频输入，始终启用思考模式", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding" + ], + "inputPrice": 6.5, + "outputPrice": 27, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "thinking": { + "default": "enabled", + "disableSupported": false, + "preserveReasoningContent": true + }, + "cacheHitPrice": 1.3, + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + } + }, + { + "modelName": "kimi-k2.7-code-highspeed", + "displayName": "Kimi K2.7 Code HighSpeed", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "月之暗面 Kimi K2.7 Code 高速版，与 K2.7 Code 参数一致，输出速度约 180 Tokens/s，短上下文可达 260 Tokens/s", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding", + "fast" + ], + "inputPrice": 13, + "outputPrice": 54, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "thinking": { + "default": "enabled", + "disableSupported": false, + "preserveReasoningContent": true + }, + "cacheHitPrice": 2.6, + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + } + }, + { + "modelName": "kimi-k2.6", + "displayName": "Kimi K2.6", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "月之暗面 Kimi K2.6，多模态模型，长程代码编写与 Agent 自主执行能力增强，256K 上下文，支持思考/非思考模式", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding" + ], + "inputPrice": 6.5, + "outputPrice": 27, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "thinking": { + "default": "enabled", + "disabledTemperature": 0.6 + }, + "cacheHitPrice": 1.1, + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + } + }, + { + "modelName": "kimi-k2.5", + "displayName": "Kimi K2.5", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "月之暗面 Kimi K2.5，原生多模态模型，256K 上下文，支持思考/非思考模式", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding" + ], + "inputPrice": 4, + "outputPrice": 21, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "thinking": { + "default": "enabled", + "disabledTemperature": 0.6 + }, + "cacheHitPrice": 0.7, + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + } + }, + { + "modelName": "moonshot-v1-8k", + "displayName": "Moonshot V1 8K", + "serviceType": [ + "chat" + ], + "description": "月之暗面 Moonshot V1 生成模型，8K 上下文", + "contextWindow": 8192, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code" + ], + "inputPrice": 2, + "outputPrice": 10, + "defaultTemperature": 0, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "moonshot-v1-32k", + "displayName": "Moonshot V1 32K", + "serviceType": [ + "chat" + ], + "description": "月之暗面 Moonshot V1 生成模型，32K 上下文", + "contextWindow": 32768, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code", + "long_context" + ], + "inputPrice": 5, + "outputPrice": 20, + "defaultTemperature": 0, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "moonshot-v1-128k", + "displayName": "Moonshot V1 128K", + "serviceType": [ + "chat" + ], + "description": "月之暗面 Moonshot V1 生成模型，128K 上下文", + "contextWindow": 131072, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code", + "long_context" + ], + "inputPrice": 10, + "outputPrice": 30, + "defaultTemperature": 0, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "moonshot-v1-8k-vision-preview", + "displayName": "Moonshot V1 8K Vision Preview", + "serviceType": [ + "vision" + ], + "description": "月之暗面 Moonshot V1 视觉模型，理解图片内容并输出文本，8K 上下文", + "contextWindow": 8192, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "image_understanding" + ], + "inputPrice": 2, + "outputPrice": 10, + "defaultTemperature": 0, + "defaultTopP": 1, + "extra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ] + } + }, + { + "modelName": "moonshot-v1-32k-vision-preview", + "displayName": "Moonshot V1 32K Vision Preview", + "serviceType": [ + "vision" + ], + "description": "月之暗面 Moonshot V1 视觉模型，理解图片内容并输出文本，32K 上下文", + "contextWindow": 32768, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "long_context" + ], + "inputPrice": 5, + "outputPrice": 20, + "defaultTemperature": 0, + "defaultTopP": 1, + "extra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ] + } + }, + { + "modelName": "moonshot-v1-128k-vision-preview", + "displayName": "Moonshot V1 128K Vision Preview", + "serviceType": [ + "vision" + ], + "description": "月之暗面 Moonshot V1 视觉模型，理解图片内容并输出文本，128K 上下文", + "contextWindow": 131072, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "long_context" + ], + "inputPrice": 10, + "outputPrice": 30, + "defaultTemperature": 0, + "defaultTopP": 1, + "extra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ] + } + } + ] +} diff --git a/compute/providers/ollama.json b/compute/providers/ollama.json new file mode 100644 index 0000000..7375415 --- /dev/null +++ b/compute/providers/ollama.json @@ -0,0 +1,34 @@ +{ + "id": "provider-local-ollama", + "provider": "ollama", + "label": "本地 Ollama", + "baseUrl": "http://localhost:11434/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat" + ], + "models": [ + { + "modelName": "llama3.1:70b", + "displayName": "Llama 3.1 70B", + "serviceType": ["chat"], + "description": "本地运行的 Llama 3.1 70B", + "contextWindow": 131072, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "code", + "reasoning", + "tool_use" + ], + "defaultTemperature": 0.8, + "defaultTopP": 0.9, + "extra": {} + } + ] +} diff --git a/compute/providers/openai-codex.json b/compute/providers/openai-codex.json new file mode 100644 index 0000000..6e5492c --- /dev/null +++ b/compute/providers/openai-codex.json @@ -0,0 +1,347 @@ +{ + "id": "provider-openai-codex-plan-001", + "provider": "openai-codex", + "brandGroup": "openai", + "label": "ChatGPT 订阅 (Codex)", + "baseUrl": "https://chatgpt.com/backend-api", + "apiFormat": "openai-codex-responses", + "apiKeyRef": "openai-codex-access-token", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "accessMode": "coding-plan", + "credentialSource": "codex-cli", + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://chatgpt.com/codex/settings/usage" + } + }, + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "gpt-6.1-sol", + "displayName": "GPT-6.1 Sol", + "serviceType": [ + "chat" + ], + "description": "GPT-6.1 Sol，面向复杂编程、电脑使用和专业工作，1.05M 上下文；工具调用使用 Responses API", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-6-sol", + "displayName": "GPT-6 Sol", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，兼顾复杂编程、智能体工作流与成本的 GPT-6 模型；1.05M 上下文", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-6-luna", + "displayName": "GPT-6 Luna", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，面向明确任务和高吞吐场景的 GPT-6 轻量模型；1.05M 上下文", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-6-astra", + "displayName": "GPT-6 Astra", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，面向复杂推理、编程和长程智能体任务，1.05M 上下文", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-reserve", + "displayName": "GPT-Reserve", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，参数暂按 GPT-5.6 Luna 配置，待官方规格确认，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "extra": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-5.6-sol", + "displayName": "GPT-5.6 Sol", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，新一代旗舰模型，复杂编程/计算机操作/网络安全能力最强，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "extra": { + "reasoning": { + "supportedEfforts": ["low", "medium", "high", "xhigh", "max"], + "defaultEffort": "low" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-5.6-terra", + "displayName": "GPT-5.6 Terra", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，均衡型模型，日常工作性能对齐 GPT-5.5 且成本更低，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "extra": { + "reasoning": { + "supportedEfforts": ["low", "medium", "high", "xhigh", "max"], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-5.6-luna", + "displayName": "GPT-5.6 Luna", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，GPT-5.6 家族中最快最省成本的档位，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "extra": { + "reasoning": { + "supportedEfforts": ["low", "medium", "high", "xhigh", "max"], + "defaultEffort": "medium" + } + }, + "source": "preset" + }, + { + "modelName": "gpt-5.5", + "displayName": "GPT-5.5", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，上一代旗舰模型，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "source": "preset" + }, + { + "modelName": "gpt-5.4", + "displayName": "GPT-5.4", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，高阶通用模型，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "source": "preset" + }, + { + "modelName": "gpt-5.4-mini", + "displayName": "GPT-5.4 Mini", + "serviceType": [ + "chat" + ], + "description": "ChatGPT 订阅额度计费，轻量高性价比模型，是 gpt-5.1-codex-mini 的官方迁移目标，272K 上下文", + "contextWindow": 272000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "source": "preset" + }, + { + "modelName": "gpt-5.3-codex-spark", + "displayName": "GPT-5.3 Codex Spark（Pro 专属）", + "serviceType": [ + "chat" + ], + "description": "ChatGPT Pro 订阅专属研究预览模型，面向实时编程迭代优化，>1000 tok/s，纯文本，128K 上下文", + "contextWindow": 128000, + "capabilities": [ + "chat", + "code", + "tool_use", + "fast" + ], + "source": "preset" + } + ], + "tombstones": [ + "gpt-5.3-codex", + "gpt-5.2-codex", + "gpt-5.1-codex-mini" + ] +} diff --git a/compute/providers/openai.json b/compute/providers/openai.json new file mode 100644 index 0000000..14a5008 --- /dev/null +++ b/compute/providers/openai.json @@ -0,0 +1,1068 @@ +{ + "id": "provider-openai-001", + "provider": "openai", + "label": "OpenAI GPT", + "baseUrl": "https://api.openai.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "openai", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "fast", + "embedding", + "tts", + "asr", + "vision", + "responses", + "image_gen", + "omni", + "realtime_voice", + "simultaneous_interpret", + "translation" + ], + "models": [ + { + "modelName": "gpt-6.1-sol", + "displayName": "GPT-6.1 Sol", + "serviceType": [ + "chat", + "responses" + ], + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "description": "GPT-6.1 Sol，面向复杂编程、电脑使用和专业工作，1.05M 上下文；工具调用使用 Responses API", + "inputPrice": 2, + "outputPrice": 10, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 2, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.1, + "reasoning": { + "supportedEfforts": ["low", "medium", "high", "xhigh", "max"], + "defaultEffort": "medium" + }, + "pricingNotes": "Prices are per 1M tokens. Cache writes $2.50/M. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request. Fast mode is 2x; Batch and Flex are 50% lower. Tool calling requires Responses API; none/minimal reasoning are unsupported.", + "responsesOnly": true + } + }, + { + "modelName": "gpt-image-2.5-sunburst", + "displayName": "GPT Image 2.5 Sunburst", + "serviceType": [ + "image_gen" + ], + "description": "图像生成与编辑模型，适合对编辑精度要求高的任务", + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision", + "high_quality" + ], + "inputPrice": 5, + "outputPrice": 30, + "extra": { + "imageInputPrice": 8, + "textCachedInputPrice": 1.25, + "imageCachedInputPrice": 2, + "pricingNotes": "Text input $5/M tokens, image input $8/M tokens, image output $30/M tokens; quality supports low, medium, high, xhigh, max, auto." + } + }, + { + "modelName": "gpt-image-2.5-flare", + "displayName": "GPT Image 2.5 Flare", + "serviceType": [ + "image_gen" + ], + "description": "快速图像生成与编辑模型，适合日常高质量出图", + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision", + "fast" + ], + "inputPrice": 5, + "outputPrice": 30, + "extra": { + "imageInputPrice": 8, + "textCachedInputPrice": 1.25, + "imageCachedInputPrice": 2, + "pricingNotes": "Text input $5/M tokens, image input $8/M tokens, image output $30/M tokens; quality supports low, medium, high, xhigh, max, auto." + } + }, + { + "modelName": "gpt-6-sol", + "displayName": "GPT-6 Sol", + "serviceType": [ + "chat", + "responses" + ], + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "description": "兼顾复杂编程、智能体工作流与成本的 GPT-6 模型；1.05M 上下文", + "inputPrice": 2, + "outputPrice": 10, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 2, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.2, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "pricingNotes": "Prices are per 1M tokens. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request. Chat Completions function calling requires reasoning_effort=none." + } + }, + { + "modelName": "gpt-6-luna", + "displayName": "GPT-6 Luna", + "serviceType": [ + "chat", + "fast", + "responses" + ], + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "description": "面向明确任务和高吞吐场景的 GPT-6 轻量模型；1.05M 上下文", + "inputPrice": 0.1, + "outputPrice": 0.5, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 3, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.01, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "pricingNotes": "Prices are per 1M tokens. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request. Chat Completions function calling requires reasoning_effort=none." + } + }, + { + "modelName": "gpt-6-astra", + "displayName": "GPT-6 Astra", + "serviceType": [ + "chat", + "responses" + ], + "description": "OpenAI GPT-6 Astra，面向复杂推理、编程和长程智能体任务，1.05M 上下文，128K 最大输出", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "inputPrice": 10, + "outputPrice": 50, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 1, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 1, + "reasoning": { + "supportedEfforts": ["low", "medium", "high", "xhigh", "max"], + "defaultEffort": "medium" + }, + "pricingNotes": "Prices are per 1M tokens. Cache writes cost $12.50. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request." + } + }, + { + "modelName": "gpt-5.5", + "displayName": "GPT-5.5", + "serviceType": [ + "chat", + "responses" + ], + "description": "OpenAI GPT-5.5 前沿模型，面向复杂推理、编码和专业工作流，1M 上下文，128K 最大输出", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "inputPrice": 5, + "outputPrice": 30, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 1, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.5, + "pricingNotes": "Prices are per 1M tokens; long-context requests may use higher-tier pricing per OpenAI pricing page.", + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ], + "defaultEffort": "medium" + } + } + }, + { + "modelName": "gpt-5.5-pro", + "displayName": "GPT-5.5 Pro", + "serviceType": [ + "responses" + ], + "description": "OpenAI GPT-5.5 Pro，最高能力推理模型，仅面向 Responses API", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "inputPrice": 30, + "outputPrice": 180, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 2, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "responsesOnly": true, + "pricingNotes": "No cached-input discount is listed for GPT-5.5 Pro on the official pricing page.", + "reasoning": { + "supportedEfforts": [ + "medium", + "high", + "xhigh" + ], + "defaultEffort": "high" + } + } + }, + { + "modelName": "gpt-5.4", + "displayName": "GPT-5.4", + "serviceType": [ + "chat", + "responses" + ], + "description": "OpenAI GPT-5.4，高性价比前沿模型，1M 上下文，支持文本和图像输入", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "inputPrice": 2.5, + "outputPrice": 15, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 3, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.25, + "pricingNotes": "Prices are per 1M tokens; long-context requests may use higher-tier pricing per OpenAI pricing page.", + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ] + } + } + }, + { + "modelName": "gpt-5.4-pro", + "displayName": "GPT-5.4 Pro", + "serviceType": [ + "responses" + ], + "description": "OpenAI GPT-5.4 Pro，专业推理与长程任务模型，仅面向 Responses API", + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "inputPrice": 30, + "outputPrice": 180, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 4, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "responsesOnly": true, + "reasoning": { + "supportedEfforts": [ + "medium", + "high", + "xhigh" + ], + "defaultEffort": "medium" + } + } + }, + { + "modelName": "gpt-5.4-mini", + "displayName": "GPT-5.4 Mini", + "serviceType": [ + "chat", + "fast", + "responses" + ], + "description": "OpenAI GPT-5.4 Mini，低延迟高性价比模型，400K 上下文，128K 最大输出", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "inputPrice": 0.75, + "outputPrice": 4.5, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 5, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.075, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ] + } + } + }, + { + "modelName": "gpt-5.4-nano", + "displayName": "GPT-5.4 Nano", + "serviceType": [ + "fast", + "responses" + ], + "description": "OpenAI GPT-5.4 Nano，最低成本极速模型，400K 上下文，128K 最大输出", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "fast", + "long_context", + "tool_use" + ], + "inputPrice": 0.2, + "outputPrice": 1.25, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 6, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + }, + "cachedInputPrice": 0.02, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ] + } + } + }, + { + "modelName": "gpt-image-2", + "displayName": "GPT Image 2", + "serviceType": [ + "image_gen" + ], + "description": "OpenAI GPT Image 2 图像生成和编辑模型，支持文本输入、图像输入与图像输出", + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision" + ], + "inputPrice": 5, + "outputPrice": 30, + "extra": { + "imageInputPrice": 8, + "textCachedInputPrice": 1.25, + "imageCachedInputPrice": 2, + "snapshot": "gpt-image-2-2026-04-21", + "pricingNotes": "Text input is $5/M tokens, image input is $8/M tokens, image output is $30/M tokens." + } + }, + { + "modelName": "gpt-realtime-2.1", + "displayName": "GPT Realtime 2.1", + "serviceType": [ + "omni", + "realtime_voice" + ], + "description": "OpenAI 最新实时多模态语音模型，支持文本、图像和音频输入输出", + "contextWindow": 128000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "realtime", + "speech_to_speech", + "reasoning", + "tool_use" + ], + "inputPrice": 4, + "outputPrice": 24, + "extra": { + "textCachedInputPrice": 0.4, + "imageInputPrice": 5, + "imageCachedInputPrice": 0.5, + "audioInputPrice": 32, + "audioCachedInputPrice": 0.4, + "audioOutputPrice": 64 + } + }, + { + "modelName": "gpt-realtime-2.1-mini", + "displayName": "GPT Realtime 2.1 Mini", + "serviceType": [ + "realtime_voice" + ], + "description": "OpenAI 低成本实时语音模型，适合实时语音交互", + "contextWindow": 128000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "realtime", + "speech_to_speech", + "fast" + ], + "inputPrice": 0.6, + "outputPrice": 2.4, + "extra": { + "textCachedInputPrice": 0.06, + "imageInputPrice": 0.8, + "imageCachedInputPrice": 0.08, + "audioInputPrice": 10, + "audioCachedInputPrice": 0.3, + "audioOutputPrice": 20 + } + }, + { + "modelName": "gpt-realtime-translate", + "displayName": "GPT Realtime Translate", + "serviceType": [ + "simultaneous_interpret", + "translation" + ], + "description": "OpenAI 实时语音翻译模型，按分钟计费", + "contextWindow": 16000, + "maxOutputTokens": 2000, + "capabilities": [ + "translation", + "simultaneous_interpretation", + "asr", + "realtime" + ], + "extra": { + "audioOutputPricePerMinute": 0.034, + "pricingNotes": "OpenAI lists this model at $0.034 per minute." + } + }, + { + "modelName": "gpt-5.2", + "displayName": "GPT-5.2", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-5.2，400K 上下文，面向编码和智能体任务", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "inputPrice": 1.75, + "outputPrice": 14, + "extra": {} + }, + { + "modelName": "gpt-5.2-pro", + "displayName": "GPT-5.2 Pro", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-5.2 Pro 专业版，最高性能", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "inputPrice": 21, + "outputPrice": 168, + "extra": {} + }, + { + "modelName": "gpt-5.1", + "displayName": "GPT-5.1", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-5.1，400K 上下文，高性能旗舰", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "inputPrice": 1.25, + "outputPrice": 10, + "extra": {} + }, + { + "modelName": "gpt-5", + "displayName": "GPT-5", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-5 新一代旗舰模型，统一所有模型能力", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "inputPrice": 1.25, + "outputPrice": 10, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 7, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + } + } + }, + { + "modelName": "gpt-5-pro", + "displayName": "GPT-5 Pro", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-5 Pro 专业版高性能推理模型", + "contextWindow": 400000, + "maxOutputTokens": 272000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "inputPrice": 15, + "outputPrice": 120, + "extra": {} + }, + { + "modelName": "gpt-5-mini", + "displayName": "GPT-5-mini", + "serviceType": [ + "chat" + ], + "description": "高性价比模型，适合日常对话，400K 上下文", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "fast", + "tool_use", + "code", + "vision" + ], + "inputPrice": 0.25, + "outputPrice": 2, + "extra": {} + }, + { + "modelName": "gpt-5-nano", + "displayName": "GPT-5-nano", + "serviceType": [ + "fast" + ], + "description": "极速响应模型，适合简单任务，400K 上下文", + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "code", + "fast", + "tool_use" + ], + "inputPrice": 0.05, + "outputPrice": 0.4, + "extra": {} + }, + { + "modelName": "gpt-4.1", + "displayName": "GPT-4.1", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-4.1，支持 1M 上下文，编码和指令遵循能力大幅提升", + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "inputPrice": 2, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 8, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + } + } + }, + { + "modelName": "gpt-4.1-mini", + "displayName": "GPT-4.1 mini", + "serviceType": [ + "chat" + ], + "description": "OpenAI GPT-4.1 mini，1M 上下文，高性价比", + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "inputPrice": 0.4, + "outputPrice": 1.6, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 9, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + } + } + }, + { + "modelName": "gpt-4.1-nano", + "displayName": "GPT-4.1 nano", + "serviceType": [ + "fast" + ], + "description": "OpenAI GPT-4.1 nano，1M 上下文，极致性价比", + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "code", + "ultra_long_context", + "fast", + "tool_use" + ], + "inputPrice": 0.1, + "outputPrice": 0.4, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "gpt-4o", + "displayName": "GPT-4o", + "serviceType": [ + "vision" + ], + "description": "多模态模型，支持图像理解", + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "tool_use" + ], + "inputPrice": 2.5, + "outputPrice": 10, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "gpt-4o-mini", + "displayName": "GPT-4o mini", + "serviceType": [ + "vision" + ], + "description": "OpenAI GPT-4o mini，高性价比多模态模型", + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "vision", + "tool_use" + ], + "inputPrice": 0.15, + "outputPrice": 0.6, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "text-embedding-3-small", + "displayName": "text-embedding-3-small", + "serviceType": [ + "embedding" + ], + "description": "轻量文本向量化模型，1536 维", + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "inputPrice": 0.02, + "extra": {} + }, + { + "modelName": "text-embedding-3-large", + "displayName": "text-embedding-3-large", + "serviceType": [ + "embedding" + ], + "description": "高精度文本向量化模型，3072 维", + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "inputPrice": 0.13, + "extra": {} + }, + { + "modelName": "tts-1", + "displayName": "TTS-1", + "serviceType": [ + "tts" + ], + "description": "标准语音合成", + "capabilities": [ + "tts" + ], + "inputPrice": 15, + "extra": { + "voices": [ + "alloy", + "echo", + "fable", + "onyx", + "nova", + "shimmer" + ], + "supportedAudioFormats": [ + "mp3", + "opus", + "aac", + "flac" + ] + } + }, + { + "modelName": "tts-1-hd", + "displayName": "TTS-1-HD", + "serviceType": [ + "tts" + ], + "description": "高清语音合成", + "capabilities": [ + "tts" + ], + "inputPrice": 30, + "extra": { + "voices": [ + "alloy", + "echo", + "fable", + "onyx", + "nova", + "shimmer" + ], + "supportedAudioFormats": [ + "mp3", + "opus", + "aac", + "flac" + ] + } + }, + { + "modelName": "whisper-1", + "displayName": "Whisper", + "serviceType": [ + "asr" + ], + "description": "通用语音识别", + "capabilities": [ + "asr", + "multilingual" + ], + "inputPrice": 0.006, + "extra": { + "maxAudioLength": 600, + "supportedInputFormats": [ + "mp3", + "mp4", + "mpeg", + "mpga", + "m4a", + "wav", + "webm" + ], + "streamingSupported": false + } + }, + { + "modelName": "o3", + "displayName": "o3", + "serviceType": [ + "responses" + ], + "description": "OpenAI o3 推理模型，200K 上下文，支持工具调用", + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "tool_use" + ], + "inputPrice": 2, + "outputPrice": 8, + "extra": {} + }, + { + "modelName": "o3-pro", + "displayName": "o3-pro", + "serviceType": [ + "responses" + ], + "description": "OpenAI o3-pro 高级推理模型，200K 上下文", + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "math", + "science" + ], + "inputPrice": 20, + "outputPrice": 80, + "extra": {} + }, + { + "modelName": "o3-mini", + "displayName": "o3-mini", + "serviceType": [ + "responses" + ], + "description": "OpenAI o3-mini 推理模型，200K 上下文", + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "math", + "code", + "fast" + ], + "inputPrice": 1.1, + "outputPrice": 4.4, + "extra": {} + }, + { + "modelName": "o4-mini", + "displayName": "o4-mini", + "serviceType": [ + "responses" + ], + "description": "OpenAI o4-mini 推理模型，200K 上下文，支持工具调用", + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "math", + "code", + "tool_use", + "vision" + ], + "inputPrice": 1.1, + "outputPrice": 4.4, + "extra": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 10, + "searchRequestPriceUsd": 0.01, + "toolType": "web_search" + } + } + } + ], + "tombstones": [ + "dall-e-3", + "gpt-4o-realtime", + "gpt-4o-realtime-preview" + ] +} diff --git a/compute/providers/openrouter.json b/compute/providers/openrouter.json new file mode 100644 index 0000000..cda6356 --- /dev/null +++ b/compute/providers/openrouter.json @@ -0,0 +1,318 @@ +{ + "id": "provider-openrouter-001", + "provider": "openrouter", + "label": "OpenRouter", + "baseUrl": "https://openrouter.ai/api/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "openrouter", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "reasoning", + "vision" + ], + "models": [ + { + "modelName": "cohere/command-a-plus", + "displayName": "Cohere: Command A+", + "serviceType": [ + "chat" + ], + "contextWindow": 192000, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 0.3, + "outputPrice": 1.5, + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准", + "maxOutputTokens": 64000 + }, + { + "modelName": "mistralai/mistral-medium-3-5", + "displayName": "Mistral: Mistral Medium 3.5", + "serviceType": [ + "chat" + ], + "contextWindow": 262144, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 1.5, + "outputPrice": 7.5, + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准", + "maxOutputTokens": 209715 + }, + { + "modelName": "google/gemini-3.8-flash", + "displayName": "Google: Gemini 3.8 Flash", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 0.75, + "outputPrice": 3.75, + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准", + "maxOutputTokens": 65536 + }, + { + "modelName": "minimax/minimax-m3", + "displayName": "MiniMax: MiniMax M3", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 0.3, + "outputPrice": 1.2, + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准", + "maxOutputTokens": 512000 + }, + { + "modelName": "qwen/qwen3.8-omni-flash", + "displayName": "Qwen: Qwen3.8 Omni Flash", + "serviceType": [ + "chat" + ], + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 0.15, + "outputPrice": 0.47, + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准", + "maxOutputTokens": 131072 + }, + { + "modelName": "qwen/qwen3.8-max-prime", + "displayName": "Qwen: Qwen3.8 Max Prime", + "serviceType": [ + "chat" + ], + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 4.0, + "outputPrice": 12.0, + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准", + "maxOutputTokens": 131072 + }, + { + "modelName": "deepseek/deepseek-v4.1-flash", + "displayName": "DeepSeek: DeepSeek V4.1 Flash", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 0.3, + "outputPrice": 1.2, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 943718 + }, + { + "modelName": "tencent/hy4-preview", + "displayName": "Tencent: Hy4 preview", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "tool_use", + "reasoning" + ], + "inputPrice": 0.834, + "outputPrice": 2.501, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 64000 + }, + { + "modelName": "z-ai/glm-5.3-flash", + "displayName": "Z.ai: GLM 5.3 Flash", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 0.15, + "outputPrice": 0.5, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 943717 + }, + { + "modelName": "z-ai/glm-5.3", + "displayName": "Z.ai: GLM 5.3", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "tool_use", + "reasoning" + ], + "inputPrice": 1.4, + "outputPrice": 4.4, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 131072 + }, + { + "modelName": "moonshotai/kimi-k3", + "displayName": "MoonshotAI: Kimi K3", + "serviceType": [ + "chat" + ], + "contextWindow": 1048576, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 2.7, + "outputPrice": 13.5, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 943718 + }, + { + "modelName": "anthropic/claude-sonnet-5.5", + "displayName": "Anthropic: Claude Sonnet 5.5", + "serviceType": [ + "chat" + ], + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 2.0, + "outputPrice": 10.0, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 128000 + }, + { + "modelName": "anthropic/claude-opus-5.5", + "displayName": "Anthropic: Claude Opus 5.5", + "serviceType": [ + "chat" + ], + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 4.0, + "outputPrice": 20.0, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 128000 + }, + { + "modelName": "openai/gpt-6.1-sol", + "displayName": "OpenAI: GPT-6.1 Sol", + "serviceType": [ + "chat" + ], + "contextWindow": 1050000, + "capabilities": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "inputPrice": 2.0, + "outputPrice": 10.0, + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准", + "extra": { + "pricingNotes": "USD per 1M tokens; OpenRouter access-specific pricing." + }, + "maxOutputTokens": 128000 + }, + { + "modelName": "openrouter/auto", + "displayName": "OpenRouter Auto", + "serviceType": [ + "chat" + ], + "description": "OpenRouter 自动路由，智能选择最优模型", + "contextWindow": 2000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "auto_routing", + "tool_use" + ], + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "pricingNotes": "OpenRouter Models API 的 prompt/completion=-1 为动态路由占位值，实际价格取决于选中的模型，不代表免费。" + } + } + ], + "tombstones": [ + "stealth/ox-alpha", + "openai/gpt-oss-120b:free", + "qwen/qwen3-coder:free" + ] +} diff --git a/compute/providers/perplexity.json b/compute/providers/perplexity.json new file mode 100644 index 0000000..c297fd9 --- /dev/null +++ b/compute/providers/perplexity.json @@ -0,0 +1,95 @@ +{ + "id": "provider-perplexity-001", + "provider": "perplexity", + "label": "Perplexity Sonar", + "baseUrl": "https://api.perplexity.ai", + "apiFormat": "openai-completions", + "apiKeyRef": "perplexity", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat" + ], + "models": [ + { + "modelName": "sonar-pro", + "displayName": "Sonar Pro", + "serviceType": [ + "chat" + ], + "description": "Perplexity 旗舰搜索增强模型，内置实时联网搜索和引用", + "contextWindow": 200000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "web_search", + "reasoning", + "citation" + ], + "inputPrice": 3.0, + "outputPrice": 15.0, + "extra": { + "requestPricingPer1k": { + "low": 6, + "medium": 10, + "high": 14 + }, + "pricingNotes": "Total Sonar API cost includes token costs plus a request fee based on search context size." + } + }, + { + "modelName": "sonar-reasoning-pro", + "displayName": "Sonar Reasoning Pro", + "serviceType": [ + "chat" + ], + "description": "Perplexity 深度推理模型，内置联网搜索和深度思考", + "contextWindow": 128000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "web_search", + "reasoning", + "deep_thinking" + ], + "inputPrice": 2.0, + "outputPrice": 8.0, + "extra": { + "requestPricingPer1k": { + "low": 6, + "medium": 10, + "high": 14 + }, + "pricingNotes": "Total Sonar API cost includes token costs plus a request fee based on search context size." + } + }, + { + "modelName": "sonar", + "displayName": "Sonar", + "serviceType": [ + "chat" + ], + "description": "Perplexity 轻量搜索模型，低成本联网搜索", + "contextWindow": 128000, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "web_search", + "citation", + "fast" + ], + "inputPrice": 1.0, + "outputPrice": 1.0, + "extra": { + "requestPricingPer1k": { + "low": 5, + "medium": 8, + "high": 12 + }, + "pricingNotes": "Total Sonar API cost includes token costs plus a request fee based on search context size." + } + } + ] +} diff --git a/compute/providers/siliconflow.json b/compute/providers/siliconflow.json new file mode 100644 index 0000000..07e5fbb --- /dev/null +++ b/compute/providers/siliconflow.json @@ -0,0 +1,80 @@ +{ + "id": "provider-siliconflow-001", + "provider": "siliconflow", + "label": "硅基流动", + "baseUrl": "https://api.siliconflow.cn/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "siliconflow", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "embedding" + ], + "models": [ + { + "modelName": "Qwen/Qwen3-Coder-480B-A35B-Instruct", + "displayName": "Qwen3 Coder 480B (硅基)", + "serviceType": [ + "chat" + ], + "description": "Qwen3-Coder-480B-A35B-Instruct，480B MoE（35B 激活），原生 256K 上下文，可通过外推扩展至 1M，面向 Agentic Coding", + "contextWindow": 262144, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "code", + "long_context", + "agent", + "tool_use" + ], + "inputPrice": 8, + "outputPrice": 16, + "defaultTemperature": 0.7, + "defaultTopP": 0.7, + "extra": {} + }, + { + "modelName": "Qwen/Qwen3-235B-A22B-Instruct-2507", + "displayName": "Qwen3 235B (硅基)", + "serviceType": [ + "chat" + ], + "description": "Qwen3-235B-A22B-Instruct-2507，235B MoE（22B 激活）通用指令模型，262K 上下文", + "contextWindow": 262144, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "code", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 2.5, + "outputPrice": 10, + "defaultTemperature": 0.7, + "defaultTopP": 0.7, + "extra": {} + }, + { + "modelName": "BAAI/bge-m3", + "displayName": "BGE-M3 (硅基)", + "serviceType": [ + "embedding" + ], + "description": "多语言 Embedding 模型，支持 8K 上下文，免费版", + "contextWindow": 8192, + "maxOutputTokens": 0, + "capabilities": [ + "text_embedding", + "multilingual" + ], + "inputPrice": 0, + "outputPrice": 0, + "extra": {} + } + ] +} diff --git a/compute/providers/stability.json b/compute/providers/stability.json new file mode 100644 index 0000000..e12b65d --- /dev/null +++ b/compute/providers/stability.json @@ -0,0 +1,43 @@ +{ + "id": "provider-stability-001", + "provider": "stability", + "label": "Stability SDXL", + "baseUrl": "https://api.stability.ai/v2beta", + "apiFormat": "stability-rest-api", + "apiKeyRef": "stability", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "image_gen" + ], + "models": [ + { + "modelName": "stable-diffusion-3.5-large", + "displayName": "Stable Diffusion 3.5 Large", + "serviceType": ["image_gen"], + "description": "Stability AI 旗舰文生图模型", + "capabilities": [ + "image_generation", + "high_quality", + "controllable" + ], + "extra": { + "supportedImageSizes": [ + "1024x1024", + "1152x896", + "896x1152", + "1216x832", + "832x1216" + ], + "supportedStyles": [ + "photographic", + "digital-art", + "anime", + "comic-book" + ] + } + } + ] +} diff --git a/compute/providers/tencent.json b/compute/providers/tencent.json new file mode 100644 index 0000000..bca88f3 --- /dev/null +++ b/compute/providers/tencent.json @@ -0,0 +1,180 @@ +{ + "id": "provider-tencent-001", + "provider": "tencent", + "label": "腾讯混元", + "baseUrl": "https://api.hunyuan.cloud.tencent.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "tencent", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "hunyuan-2.0-thinking-20251109", + "displayName": "混元 2.0 Think", + "serviceType": [ + "reasoning" + ], + "description": "腾讯混元最新推理模型，128K输入/64K输出，MoE架构406B总参数", + "contextWindow": 196608, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "long_context" + ], + "inputPrice": 3.975, + "outputPrice": 15.9, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32768, + "inputPrice": 3.975, + "outputPrice": 15.9 + }, + { + "maxInputTokens": 131072, + "inputPrice": 5.3, + "outputPrice": 21.2 + } + ] + } + }, + { + "modelName": "hunyuan-2.0-instruct-20251111", + "displayName": "混元 2.0 Instruct", + "serviceType": [ + "chat" + ], + "description": "腾讯混元 2.0 指令模型，最大输入128K/最大输出16K，支持联网搜索和 Function Calling", + "contextWindow": 147456, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "long_context", + "tool_use" + ], + "inputPrice": 3.18, + "outputPrice": 7.95, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32768, + "inputPrice": 3.18, + "outputPrice": 7.95 + }, + { + "maxInputTokens": 131072, + "inputPrice": 4.505, + "outputPrice": 11.13 + } + ] + } + }, + { + "modelName": "hunyuan-turbos-latest", + "displayName": "混元 Turbo S", + "serviceType": [ + "chat" + ], + "description": "腾讯混元高速模型，32K输入/16K输出，支持推理和工具调用", + "contextWindow": 32768, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "fast", + "tool_use" + ], + "inputPrice": 0.8, + "outputPrice": 2, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "hunyuan-t1-latest", + "displayName": "混元 T1", + "serviceType": [ + "reasoning" + ], + "description": "腾讯混元 T1 深度推理模型，256K 上下文，专攻数学/代码/复杂推理", + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "long_context" + ], + "inputPrice": 5, + "outputPrice": 20, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "hunyuan-t1-vision", + "displayName": "混元 T1 Vision", + "serviceType": [ + "reasoning" + ], + "description": "腾讯混元 T1 视觉推理模型，支持图片理解 + 深度推理", + "contextWindow": 131072, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "vision", + "image_understanding" + ], + "inputPrice": 5, + "outputPrice": 20, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "hunyuan-turbos-vision", + "displayName": "混元 Turbo S Vision", + "serviceType": [ + "chat" + ], + "description": "腾讯混元 Turbo S 视觉模型，支持图片理解，高速推理", + "contextWindow": 32768, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "image_understanding", + "fast", + "tool_use" + ], + "inputPrice": 0.8, + "outputPrice": 2, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + } + ] +} diff --git a/compute/providers/volcengine.json b/compute/providers/volcengine.json new file mode 100644 index 0000000..b87e9ce --- /dev/null +++ b/compute/providers/volcengine.json @@ -0,0 +1,712 @@ +{ + "id": "provider-volcengine-001", + "provider": "volcengine", + "label": "火山 豆包", + "baseUrl": "https://ark.cn-beijing.volces.com/api/v3", + "apiFormat": "openai-completions", + "apiKeyRef": "volcengine", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning", + "vision", + "embedding", + "voice_clone", + "realtime_voice", + "simultaneous_interpret", + "translation", + "image_gen", + "video_gen" + ], + "models": [ + { + "modelName": "doubao-seed-evolving", + "displayName": "Doubao Seed Evolving", + "contextWindow": 1000000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "description": "Seed 系列持续迭代模型，通过固定 ID 获取最新版本，2026-09-09 更新" + }, + { + "modelName": "doubao-seed-2.1-pro", + "displayName": "豆包 Seed 2.1 Pro", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "火山方舟豆包 Seed 2.1 Pro，生产级 Coding、Agent 与多模态旗舰模型", + "contextWindow": 256000, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "long_context", + "tool_use", + "agent", + "deep_thinking" + ], + "inputPrice": 6, + "outputPrice": 30, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "apiModelId": "doubao-seed-2-1-pro-260628", + "extra": { + "cacheHitPrice": 1.2, + "cacheStoragePricePerMillionTokenHour": 0.017, + "supportsResponsesApi": true, + "supportsChatApi": true + } + }, + { + "modelName": "doubao-seed-2.1-turbo", + "displayName": "豆包 Seed 2.1 Turbo", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "火山方舟豆包 Seed 2.1 Turbo，高吞吐低成本生产模型，适合规模化 Agent 和 Coding 调用", + "contextWindow": 256000, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "long_context", + "tool_use", + "agent", + "fast" + ], + "inputPrice": 3, + "outputPrice": 15, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "apiModelId": "doubao-seed-2-1-turbo-260628", + "extra": { + "cacheHitPrice": 0.6, + "cacheStoragePricePerMillionTokenHour": 0.017, + "supportsResponsesApi": true, + "supportsChatApi": true + } + }, + { + "modelName": "doubao-seed-2.0-pro", + "displayName": "豆包 2.0 Pro", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包 Seed 2.0 旗舰模型，256K上下文，支持复杂推理、Agent 和多模态理解", + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "inputPrice": 3.2, + "outputPrice": 16, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 3.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 128000, + "inputPrice": 4.8, + "outputPrice": 24 + }, + { + "maxInputTokens": 256000, + "inputPrice": 9.6, + "outputPrice": 48 + } + ] + } + }, + { + "modelName": "doubao-seed-2.0-lite", + "displayName": "豆包 2.0 Lite", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包 Seed 2.0 全模态均衡模型，256K上下文，支持视频/图像/音频/文本统一理解", + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "audio_understanding", + "video_understanding" + ], + "inputPrice": 0.6, + "outputPrice": 3.6, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.6, + "outputPrice": 3.6 + }, + { + "maxInputTokens": 128000, + "inputPrice": 0.9, + "outputPrice": 5.4 + }, + { + "maxInputTokens": 256000, + "inputPrice": 1.8, + "outputPrice": 10.8 + } + ] + } + }, + { + "modelName": "doubao-seed-2.0-mini", + "displayName": "豆包 2.0 Mini", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包 Seed 2.0 全模态轻量模型，256K上下文，低延迟高并发，支持全模态理解", + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "audio_understanding", + "video_understanding", + "fast" + ], + "inputPrice": 0.2, + "outputPrice": 2, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.2, + "outputPrice": 2 + }, + { + "maxInputTokens": 128000, + "inputPrice": 0.4, + "outputPrice": 4 + }, + { + "maxInputTokens": 256000, + "inputPrice": 0.8, + "outputPrice": 8 + } + ] + } + }, + { + "modelName": "doubao-seed-2.0-code", + "displayName": "豆包 2.0 Code", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包 Seed 2.0 编程模型，面向 Agentic Coding 场景优化", + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "long_context", + "tool_use", + "vision", + "deep_thinking" + ], + "inputPrice": 3.2, + "outputPrice": 16, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 3.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 128000, + "inputPrice": 4.8, + "outputPrice": 24 + }, + { + "maxInputTokens": 256000, + "inputPrice": 9.6, + "outputPrice": 48 + } + ] + } + }, + { + "modelName": "doubao-seed-1.8", + "displayName": "豆包 Seed-1.8", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包1.8旗舰模型，256K上下文，支持深度思考和多模态", + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "inputPrice": 0.8, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.8, + "outputPrice": 8, + "pricingNotes": "Short outputs up to 0.2K tokens are priced at 2 CNY/M output tokens." + }, + { + "maxInputTokens": 128000, + "inputPrice": 1.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 256000, + "inputPrice": 2.4, + "outputPrice": 24 + } + ], + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + } + }, + { + "modelName": "doubao-seed-1.6", + "displayName": "豆包 Seed-1.6", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包1.6旗舰模型，256K上下文", + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "inputPrice": 0.8, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.8, + "outputPrice": 8, + "pricingNotes": "Short outputs up to 0.2K tokens are priced at 2 CNY/M output tokens." + }, + { + "maxInputTokens": 128000, + "inputPrice": 1.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 256000, + "inputPrice": 2.4, + "outputPrice": 24 + } + ], + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + } + }, + { + "modelName": "doubao-seed-1.6-thinking", + "displayName": "豆包 Seed-1.6 思考版", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包1.6深度思考模型，256K上下文", + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "inputPrice": 0.8, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + } + }, + { + "modelName": "doubao-seed-1.6-flash", + "displayName": "豆包 Seed-1.6 Flash", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包1.6快速版，成本较前代下降63%", + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "chat", + "code", + "fast", + "long_context", + "vision" + ], + "inputPrice": 0.15, + "outputPrice": 1.5, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + } + }, + { + "modelName": "doubao-seed-1.6-lite", + "displayName": "豆包 Seed-1.6 Lite", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包1.6轻量版，极致性价比", + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "code", + "fast" + ], + "inputPrice": 0.3, + "outputPrice": 2.4, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "pricingTiers": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.3, + "outputPrice": 2.4 + }, + { + "maxInputTokens": 128000, + "inputPrice": 0.6, + "outputPrice": 4 + }, + { + "maxInputTokens": 256000, + "inputPrice": 1.2, + "outputPrice": 12 + } + ], + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + } + }, + { + "modelName": "doubao-seed-1.6-vision", + "displayName": "豆包 Seed-1.6 视觉版", + "serviceType": [ + "vision" + ], + "description": "火山引擎豆包1.6视觉模型，256K上下文，支持视频理解", + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "gui_agent", + "long_context" + ], + "inputPrice": 0.8, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": { + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + } + }, + { + "modelName": "doubao-seed-code", + "displayName": "豆包 Seed-Code", + "serviceType": [ + "chat" + ], + "description": "火山引擎豆包编程专用模型，256K上下文", + "contextWindow": 256000, + "maxOutputTokens": 32768, + "capabilities": [ + "code", + "reasoning", + "long_context", + "deep_thinking" + ], + "inputPrice": 1.2, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.7, + "extra": {} + }, + { + "modelName": "deepseek-v3.2", + "displayName": "DeepSeek V3.2 (火山引擎)", + "serviceType": [ + "chat" + ], + "description": "火山引擎托管的DeepSeek V3.2，128K上下文", + "contextWindow": 128000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "inputPrice": 2, + "outputPrice": 3, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": {} + }, + { + "modelName": "deepseek-r1", + "displayName": "DeepSeek R1 (火山引擎)", + "serviceType": [ + "reasoning" + ], + "description": "火山引擎托管的DeepSeek R1推理模型，128K上下文", + "contextWindow": 128000, + "maxOutputTokens": 65536, + "capabilities": [ + "reasoning", + "math", + "code", + "science", + "deep_thinking" + ], + "inputPrice": 4, + "outputPrice": 16, + "extra": {} + }, + { + "modelName": "doubao-embedding", + "displayName": "豆包 Embedding", + "serviceType": [ + "embedding" + ], + "description": "火山引擎豆包标准向量模型，默认2048维度，可选512/1024/2048维度", + "contextWindow": 4096, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "chinese_optimized" + ], + "inputPrice": 0.5, + "extra": {} + }, + { + "modelName": "doubao-embedding-large", + "displayName": "豆包 Embedding Large", + "serviceType": [ + "embedding" + ], + "description": "火山引擎豆包大号向量模型，默认2048维度，可选512/1024/2048/4096维度", + "contextWindow": 4096, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "chinese_optimized" + ], + "inputPrice": 0.5, + "extra": {} + }, + { + "modelName": "volc-mega-tts-clone", + "displayName": "火山声音复刻", + "serviceType": [ + "voice_clone" + ], + "description": "火山引擎声音复刻，支持极少样本", + "capabilities": [ + "voice_clone", + "chinese_optimized", + "few_shot" + ], + "extra": { + "maxCloneSamples": 3, + "minCloneDuration": 5 + } + }, + { + "modelName": "volc-realtime-voice", + "displayName": "火山实时语音", + "serviceType": [ + "realtime_voice" + ], + "description": "火山引擎实时语音交互", + "capabilities": [ + "realtime", + "speech_to_speech", + "low_latency", + "chinese_optimized" + ], + "extra": { + "supportedModes": [ + "speech-to-speech" + ], + "latencyMs": 200 + } + }, + { + "modelName": "volc-simultaneous", + "displayName": "火山同声传译", + "serviceType": [ + "simultaneous_interpret" + ], + "description": "火山引擎同声传译，支持中英双向", + "capabilities": [ + "simultaneous_interpretation", + "chinese_optimized", + "low_latency" + ], + "extra": { + "streamingSupported": true + } + }, + { + "modelName": "volc-translation", + "displayName": "火山翻译", + "serviceType": [ + "translation" + ], + "description": "火山引擎机器翻译，支持多语种互译", + "capabilities": [ + "translation", + "multilingual", + "chinese_optimized" + ], + "extra": {} + }, + { + "modelName": "doubao-seedream-4-0-250828", + "displayName": "豆包 Seedream 4.0（文生图/图生图）", + "serviceType": [ + "image_gen" + ], + "description": "字节 Seedream 图像模型，支持文生图、图生图（img2img）、局部重绘与多图融合（最多 10 张参考图）；OpenAI 风格 /images/generations 同步返回", + "capabilities": [ + "image_generation", + "image_to_image", + "multi_image_fusion", + "chinese_optimized", + "high_quality" + ], + "extra": { + "endpoint": "/images/generations", + "supportedSizes": [ + "1K", + "2K", + "4K" + ], + "maxFusionImages": 10 + } + }, + { + "modelName": "doubao-seedance-2-0-260128", + "displayName": "豆包 Seedance 2.0（文生/图生视频）", + "serviceType": [ + "video_gen" + ], + "description": "字节 Seedance 视频模型，支持文生视频、图生视频、多图参考、首尾帧、参考视频/音频（多模态），异步任务", + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "first_last_frame", + "reference_video", + "reference_audio", + "chinese_optimized", + "high_quality" + ], + "extra": { + "endpoint": "/contents/generations/tasks", + "maxVideoDuration": 15, + "supportedResolutions": [ + "720p", + "1080p" + ], + "maxReferenceImages": 9, + "asyncMode": true + } + } + ] +} diff --git a/compute/providers/xai.json b/compute/providers/xai.json new file mode 100644 index 0000000..3e8be38 --- /dev/null +++ b/compute/providers/xai.json @@ -0,0 +1,143 @@ +{ + "id": "provider-xai-001", + "provider": "xai", + "label": "xAI Grok", + "baseUrl": "https://api.x.ai/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "xai", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "grok-4.7", + "displayName": "Grok 4.7", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "xAI 面向编程与智能体任务的旗舰模型，500K 上下文，支持图像输入和可配置推理", + "contextWindow": 500000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "structured_output", + "long_context" + ], + "inputPrice": 2, + "outputPrice": 6, + "extra": { + "cachedInputPrice": 0.5, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh" + ], + "defaultEffort": "high" + }, + "pricingNotes": "Prices are per 1M tokens below 200K prompt tokens; above 200K, input and cache rates are 2x and output rates are 2x. No text output limit is documented." + } + }, + { + "modelName": "grok-4.3", + "displayName": "Grok 4.3", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "xAI Grok 4.3 通用模型，1M 上下文，支持可配置推理、函数调用和结构化输出", + "contextWindow": 1000000, + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "structured_output", + "long_context" + ], + "inputPrice": 1.25, + "outputPrice": 2.5, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cachedInputPrice": 0.2, + "aliases": [ + "grok-4.3-latest", + "grok-latest" + ], + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high" + ], + "defaultEffort": "low" + } + } + }, + { + "modelName": "grok-build-0.1", + "displayName": "Grok Build 0.1", + "serviceType": [ + "chat" + ], + "description": "xAI 面向 agentic coding 工作流的专用代码模型", + "contextWindow": 256000, + "capabilities": [ + "chat", + "code", + "agent", + "tool_use", + "structured_output" + ], + "inputPrice": 1, + "outputPrice": 2, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cachedInputPrice": 0.2 + } + }, + { + "modelName": "grok-4.20-0309-reasoning", + "displayName": "Grok 4.20 Reasoning (2026-03-09)", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "xAI Grok 4.20 固定版本推理模型，当前价格与 Grok 4.3 一致", + "contextWindow": 1000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "structured_output", + "long_context" + ], + "inputPrice": 1.25, + "outputPrice": 2.5, + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "cachedInputPrice": 0.2, + "legacy": true + } + } + ], + "tombstones": [ + "grok-4-1-fast-reasoning" + ] +} diff --git a/compute/providers/xiaomi.json b/compute/providers/xiaomi.json new file mode 100644 index 0000000..6924abb --- /dev/null +++ b/compute/providers/xiaomi.json @@ -0,0 +1,248 @@ +{ + "id": "provider-xiaomi-001", + "provider": "xiaomi", + "label": "小米 MiMo", + "baseUrl": "https://api.xiaomimimo.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "xiaomi", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "brandGroup": "xiaomi", + "priceCurrency": "CNY", + "accessMode": "api", + "services": [ + "chat", + "reasoning", + "vision", + "tts", + "asr" + ], + "models": [ + { + "modelName": "mimo-v2.6-pro", + "displayName": "MiMo-V2.6-Pro", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "小米全模态旗舰推理模型，适合复杂项目与长程智能体任务", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context" + ], + "inputPrice": 3, + "outputPrice": 6, + "extra": { + "cachedInputPrice": 0.025, + "thinkingDefault": true, + "pricingNotes": "CNY per 1M tokens on the Xiaomi MiMo API." + } + }, + { + "modelName": "mimo-v2.6-flash", + "displayName": "MiMo-V2.6-Flash", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "小米全模态高效推理模型，适合高频调用与规模化任务", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context", + "fast" + ], + "inputPrice": 1, + "outputPrice": 2, + "extra": { + "cachedInputPrice": 0.02, + "thinkingDefault": true, + "pricingNotes": "CNY per 1M tokens on the Xiaomi MiMo API." + } + }, + { + "modelName": "mimo-v2.6-pro-ultraspeed", + "displayName": "MiMo-V2.6-Pro-UltraSpeed", + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "MiMo-V2.6-Pro 的超高速服务，需向小米申请定制配额", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context", + "fast" + ], + "inputPrice": 30, + "outputPrice": 60, + "extra": { + "cachedInputPrice": 0.25, + "thinkingDefault": true, + "pricingNotes": "CNY per 1M tokens on the Xiaomi MiMo API; UltraSpeed requires custom rate limits." + } + }, + { + "modelName": "mimo-x-flash-preview", + "displayName": "MiMo-X-Flash Preview", + "serviceType": [ + "chat", + "vision" + ], + "description": "小米 MiMo-X Flash 内测模型，1M 上下文，支持图片识别", + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision" + ], + "extra": {} + }, + { + "modelName": "mimo-x-pro-preview", + "displayName": "MiMo-X-Pro Preview", + "serviceType": [ + "chat", + "vision" + ], + "description": "小米 MiMo-X Pro 内测模型，1M 上下文，支持图片识别", + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision" + ], + "extra": {} + }, + { + "modelName": "mimo-v2.5-pro", + "displayName": "MiMo-V2.5-Pro", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "小米 MiMo-V2.5-Pro，1M 上下文，1T MoE 参数，计划于 2026-10-21 下线", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code" + ], + "extra": { + "cachedInputPrice": 0.025 + }, + "inputPrice": 3, + "outputPrice": 6 + }, + { + "modelName": "mimo-v2.5-tts", + "displayName": "MiMo-V2.5-TTS", + "serviceType": [ + "tts" + ], + "description": "小米最新高品质语音合成模型，内置多种精调音色，支持情感/语速/语调精细控制、方言和歌唱", + "capabilities": [ + "tts", + "style_control", + "chinese_optimized" + ], + "extra": { + "voices": [ + {"id": "mimo_default", "label": "MiMo 默认"}, + {"id": "冰糖", "label": "冰糖"}, + {"id": "茉莉", "label": "茉莉"}, + {"id": "苏打", "label": "苏打"}, + {"id": "白桦", "label": "白桦"}, + {"id": "Mia", "label": "Mia"}, + {"id": "Chloe", "label": "Chloe"}, + {"id": "Milo", "label": "Milo"}, + {"id": "Dean", "label": "Dean"} + ], + "audioFormats": ["wav", "mp3", "pcm16le"], + "sampleRate": 24000, + "endpoint": "/chat/completions", + "authHeader": "api-key" + } + }, + { + "modelName": "mimo-v2.5-asr", + "displayName": "MiMo-V2.5-ASR", + "serviceType": [ + "asr" + ], + "description": "小米 MiMo 语音识别模型，支持中英文提示与整段 WAV/MP3 识别", + "contextWindow": 8192, + "maxOutputTokens": 2048, + "capabilities": [ + "asr", + "multilingual" + ], + "extra": {} + }, + { + "modelName": "mimo-v2.5-tts-voicedesign", + "displayName": "MiMo-V2.5-TTS-VoiceDesign", + "serviceType": [ + "tts" + ], + "description": "语音设计模型，通过一句话描述即可生成全新音色", + "capabilities": [ + "tts", + "voice_design" + ], + "extra": { + "audioFormats": ["mp3", "wav"], + "sampleRate": 24000, + "endpoint": "/chat/completions", + "authHeader": "api-key" + } + }, + { + "modelName": "mimo-v2.5-tts-voiceclone", + "displayName": "MiMo-V2.5-TTS-VoiceClone", + "serviceType": [ + "tts" + ], + "description": "语音克隆模型，少量样本即可高保真复刻目标音色", + "capabilities": [ + "tts", + "voice_clone" + ], + "extra": { + "audioFormats": ["mp3", "wav"], + "sampleRate": 24000, + "endpoint": "/chat/completions", + "authHeader": "api-key" + } + } + ] +} diff --git a/compute/providers/xunfei-x2.json b/compute/providers/xunfei-x2.json new file mode 100644 index 0000000..1b374c7 --- /dev/null +++ b/compute/providers/xunfei-x2.json @@ -0,0 +1,41 @@ +{ + "id": "provider-xunfei-x2-001", + "provider": "xunfei", + "label": "讯飞星火 X2", + "baseUrl": "https://spark-api-open.xf-yun.com/x2", + "apiFormat": "openai-completions", + "apiKeyRef": "xunfei", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "spark-x", + "displayName": "讯飞星火 X2", + "serviceType": [ + "reasoning" + ], + "description": "讯飞深度推理模型（X2），输入64K/输出128K，支持思考模式与函数调用", + "contextWindow": 65536, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "math", + "code", + "tool_use" + ], + "defaultTemperature": 1.2, + "defaultTopP": 0.95, + "extra": { + "pricingNote": "官方公开文档说明模型价格以控制台及实际购买页为准，未公开固定 token 单价" + } + } + ] +} diff --git a/compute/providers/xunfei.json b/compute/providers/xunfei.json new file mode 100644 index 0000000..f609f3c --- /dev/null +++ b/compute/providers/xunfei.json @@ -0,0 +1,42 @@ +{ + "id": "provider-xunfei-001", + "provider": "xunfei", + "label": "讯飞星火", + "baseUrl": "https://spark-api-open.xf-yun.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "xunfei", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ], + "models": [ + { + "modelName": "4.0Ultra", + "displayName": "讯飞星火 4.0 Ultra", + "serviceType": [ + "chat" + ], + "description": "讯飞星火最强通用模型，支持推理、代码和工具调用", + "contextWindow": 32768, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "defaultTemperature": 1, + "defaultTopP": 1, + "extra": { + "pricingNote": "官方公开文档说明模型价格以控制台及实际购买页为准，未公开固定 token 单价" + } + } + ], + "tombstones": [ + "spark-x" + ] +} diff --git a/compute/providers/zhipu-embedding.json b/compute/providers/zhipu-embedding.json new file mode 100644 index 0000000..80fc203 --- /dev/null +++ b/compute/providers/zhipu-embedding.json @@ -0,0 +1,39 @@ +{ + "id": "provider-zhipu-embedding-001", + "provider": "zhipu-embedding", + "brandGroup": "zhipu", + "label": "智谱 Embedding", + "baseUrl": "https://open.bigmodel.cn/api/paas/v4", + "apiFormat": "openai-completions", + "apiKeyRef": "zhipu", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "embedding" + ], + "models": [ + { + "modelName": "embedding-3", + "displayName": "智谱 embedding-3", + "serviceType": [ + "embedding" + ], + "description": "智谱嵌入模型v3，支持自定义维度；模型上下文窗口 8K，单条输入最多 3072 tokens。走 OpenAI 兼容 /embeddings 端点（Anthropic 协议不提供 embeddings）", + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "custom_dimensions" + ], + "inputPrice": 0.5, + "extra": { + "maxInputTokensPerItem": 3072, + "maxBatchItems": 64, + "vectorDimensions": "256-2048" + } + } + ] +} diff --git a/compute/providers/zhipu.json b/compute/providers/zhipu.json new file mode 100644 index 0000000..20245c4 --- /dev/null +++ b/compute/providers/zhipu.json @@ -0,0 +1,556 @@ +{ + "id": "provider-zhipu-001", + "provider": "zhipu", + "label": "智谱 GLM", + "baseUrl": "https://open.bigmodel.cn/api/anthropic", + "apiFormat": "anthropic-messages", + "apiKeyRef": "zhipu", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "vision", + "reasoning", + "fast" + ], + "tombstones": [ + "embedding-3", + "glm-4.7-thinking", + "glm-z1-air", + "glm-z1-airx", + "glm-z1-flash", + "glm-z1-flashx", + "glm-4-0520" + ], + "models": [ + { + "modelName": "glm-5.3-flashx", + "displayName": "GLM-5.3-FlashX", + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "video_understanding", + "tool_use", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "Z.ai GLM-5.3-FlashX，高速多模态推理模型，100 万上下文，始终开启推理", + "extra": { + "adaptiveThinking": true, + "thinkingOnly": true, + "cacheHitPrice": 0.57, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + }, + "pricingNotes": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。" + }, + "inputPrice": 2, + "outputPrice": 7 + }, + { + "modelName": "glm-5.3-flash", + "displayName": "GLM-5.3-Flash", + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "image_understanding", + "video_understanding" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "description": "智谱 GLM-5.3-Flash 原生多模态模型，混合稀疏与线性注意力，面向高效编码与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "extra": { + "adaptiveThinking": true, + "thinkingDefault": true, + "thinkingOnly": true, + "cacheHitPrice": 0.23, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + }, + "pricingNotes": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。" + }, + "inputPrice": 0.8, + "outputPrice": 2.8 + }, + { + "modelName": "glm-5.3", + "displayName": "GLM-5.3", + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "extra": { + "adaptiveThinking": true, + "thinkingDefault": true, + "thinkingOnly": true, + "cacheHitPrice": 2, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + }, + "pricingNotes": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。" + }, + "inputPrice": 8, + "outputPrice": 28 + }, + { + "modelName": "glm-5.2", + "displayName": "GLM-5.2", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-5.2 最新旗舰模型，面向复杂长程任务与项目级工程交付，1M 上下文，128K 最大输出", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "inputPrice": 8, + "outputPrice": 28, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 2, + "cacheStoragePrice": 0, + "thinking": { + "default": "enabled" + }, + "reasoning": { + "supportedEfforts": [ + "max", + "xhigh", + "high", + "medium", + "low", + "minimal", + "none" + ], + "defaultEffort": "max" + } + } + }, + { + "modelName": "glm-5.1", + "displayName": "GLM-5.1", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-5.1 旗舰模型，面向长程 Agentic Coding 与 Autonomous Agent 场景，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "inputPrice": 6, + "outputPrice": 24, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 1.3, + "cacheStoragePrice": 0, + "thinking": { + "default": "enabled" + }, + "pricingTiers": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 6, + "outputPrice": 24, + "cacheHitPrice": 1.3 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 8, + "outputPrice": 28, + "cacheHitPrice": 2 + } + ] + } + }, + { + "modelName": "glm-5-turbo", + "displayName": "GLM-5-Turbo", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-5-Turbo，高阶快速模型，面向长链路任务优化，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "fast" + ], + "inputPrice": 5, + "outputPrice": 22, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 1.2, + "cacheStoragePrice": 0, + "decodeSpeed": "40-60 tokens/s", + "thinking": { + "default": "enabled" + }, + "pricingTiers": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 5, + "outputPrice": 22, + "cacheHitPrice": 1.2 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 7, + "outputPrice": 26, + "cacheHitPrice": 1.8 + } + ] + } + }, + { + "modelName": "glm-5", + "displayName": "GLM-5", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-5，面向 Agentic Engineering 的旗舰基座模型，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "inputPrice": 4, + "outputPrice": 18, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 1, + "cacheStoragePrice": 0, + "decodeSpeed": "30-50 tokens/s", + "thinking": { + "default": "enabled" + }, + "pricingTiers": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 4, + "outputPrice": 18, + "cacheHitPrice": 1 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 6, + "outputPrice": 22, + "cacheHitPrice": 1.5 + } + ] + } + }, + { + "modelName": "glm-4.7", + "displayName": "GLM-4.7", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-4.7，高智能模型，通用对话、推理与智能体能力全面升级，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "inputPrice": 2, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 0.4, + "cacheStoragePrice": 0, + "decodeSpeed": "30-50 tokens/s", + "thinking": { + "default": "enabled", + "mode": "forced" + }, + "pricingTiers": [ + { + "condition": "input_length_k_tokens: [0, 32); output_length_k_tokens: [0, 0.2)", + "inputPrice": 2, + "outputPrice": 8, + "cacheHitPrice": 0.4 + }, + { + "condition": "input_length_k_tokens: [0, 32); output_length_k_tokens: [0.2+)", + "inputPrice": 3, + "outputPrice": 14, + "cacheHitPrice": 0.6 + }, + { + "condition": "input_length_k_tokens: [32, 200)", + "inputPrice": 4, + "outputPrice": 16, + "cacheHitPrice": 0.8 + } + ] + } + }, + { + "modelName": "glm-4.7-flashx", + "displayName": "GLM-4.7-FlashX", + "serviceType": [ + "chat", + "fast" + ], + "description": "智谱 GLM-4.7-FlashX 轻量高速模型，适用于中文写作、翻译、角色扮演等通用场景，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "fast", + "long_context", + "multilingual", + "tool_use" + ], + "inputPrice": 0.5, + "outputPrice": 3, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 0.1, + "cacheStoragePrice": 0, + "decodeSpeed": "30-40 tokens/s" + } + }, + { + "modelName": "glm-5v-turbo", + "displayName": "GLM-5V-Turbo", + "serviceType": [ + "vision" + ], + "description": "智谱首个多模态 Agent 基座模型，支持视频、图像、文本和文件输入，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "file_understanding", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "inputPrice": 5, + "outputPrice": 22, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 1.2, + "cacheStoragePrice": 0, + "decodeSpeed": "30-50 tokens/s", + "thinking": { + "default": "enabled" + }, + "inputModalities": [ + "image", + "video", + "file", + "text" + ], + "pricingTiers": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 5, + "outputPrice": 22, + "cacheHitPrice": 1.2 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 7, + "outputPrice": 26, + "cacheHitPrice": 1.8 + } + ] + } + }, + { + "modelName": "glm-4.6v", + "displayName": "GLM-4.6V", + "serviceType": [ + "vision" + ], + "description": "智谱 GLM-4.6V 多模态模型，支持图像、视频、文本和文件理解，128K 上下文，32K 最大输出", + "contextWindow": 128000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "file_understanding", + "long_context", + "tool_use", + "reasoning" + ], + "inputPrice": 1, + "outputPrice": 3, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 0.2, + "cacheStoragePrice": 0, + "decodeSpeed": "30-50 tokens/s", + "thinking": { + "default": "enabled" + }, + "inputModalities": [ + "image", + "video", + "file", + "text" + ], + "pricingTiers": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 1, + "outputPrice": 3, + "cacheHitPrice": 0.2 + }, + { + "condition": "input_length_k_tokens: [32, 128)", + "inputPrice": 2, + "outputPrice": 6, + "cacheHitPrice": 0.4 + } + ] + } + }, + { + "modelName": "glm-4.6", + "displayName": "GLM-4.6", + "serviceType": [ + "chat", + "reasoning" + ], + "description": "智谱 GLM-4.6，擅长高级编码、复杂推理与工具调用，200K 上下文，128K 最大输出", + "contextWindow": 200000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "tool_use" + ], + "inputPrice": 2, + "outputPrice": 8, + "defaultTemperature": 1, + "defaultTopP": 0.95, + "extra": { + "cacheHitPrice": 0.5, + "thinking": { + "default": "enabled" + } + } + } + ] +} diff --git a/compute/service-map.json b/compute/service-map.json new file mode 100644 index 0000000..70ee24b --- /dev/null +++ b/compute/service-map.json @@ -0,0 +1,82 @@ +{ + "chat": { + "modelName": "gpt-5.4-mini", + "providerId": "provider-openai-001" + }, + "reasoning": { + "modelName": "deepseek-flash", + "providerId": "provider-deepseek-001" + }, + "fast": { + "modelName": "gpt-5.4-nano", + "providerId": "provider-openai-001" + }, + "embedding": { + "modelName": "text-embedding-3-small", + "providerId": "provider-openai-001" + }, + "rerank": { + "modelName": "qwen3-rerank", + "providerId": "provider-dashscope-001" + }, + "vision": { + "modelName": "gpt-5.4-mini", + "providerId": "provider-openai-001" + }, + "tts": { + "modelName": "tts-1", + "providerId": "provider-openai-001" + }, + "asr": { + "modelName": "whisper-large-v3", + "providerId": "provider-local-whisper" + }, + "ocr": { + "modelName": "qwen3-vl-plus", + "providerId": "provider-dashscope-001" + }, + "responses": { + "modelName": "gpt-5.5", + "providerId": "provider-openai-001" + }, + "image_gen": { + "modelName": "image-01", + "providerId": "provider-minimax-001" + }, + "video_gen": { + "modelName": "MiniMax-Hailuo-2.3", + "providerId": "provider-minimax-001" + }, + "omni": { + "modelName": "gpt-realtime-2.1", + "providerId": "provider-openai-001" + }, + "voice_clone": { + "modelName": "cosyvoice-clone", + "providerId": "provider-dashscope-001" + }, + "realtime_voice": { + "modelName": "gpt-realtime-2.1", + "providerId": "provider-openai-001" + }, + "simultaneous_interpret": { + "modelName": "volc-simultaneous", + "providerId": "provider-volcengine-001" + }, + "translation": { + "modelName": "qwen-mt-plus", + "providerId": "provider-dashscope-001" + }, + "computer_use": { + "modelName": "claude-sonnet-5", + "providerId": "provider-anthropic-001" + }, + "music_gen": { + "modelName": "music-2.6", + "providerId": "provider-minimax-001" + }, + "summary": { + "modelName": "mimo-v2.6-flash", + "providerId": "desirecore-cloud" + } +} diff --git a/compute/v2/catalog.json b/compute/v2/catalog.json new file mode 100644 index 0000000..49904dd --- /dev/null +++ b/compute/v2/catalog.json @@ -0,0 +1,30343 @@ +{ + "formatVersion": 2, + "publication": { + "state": "candidate", + "requiredClientVersion": null + }, + "specs": [ + { + "modelRef": "anthropic:claude-opus-5-5", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-opus-5-5" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Opus 5.5，始终开启自适应思考，默认 medium effort", + "constraints": { + "thinkingOnly": true, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 22, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json", + "recordedFacts": { + "releasedAt": "2026-09-22" + } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-fable-5-1", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-fable-5-1" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Fable 5.1，面向高难推理和长程智能体任务，始终开启自适应思考", + "constraints": { + "thinkingOnly": true, + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-opus-5", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-opus-5" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Opus 5，Anthropic Opus 系列当前旗舰；思考默认开启，effort 支持到 max", + "constraints": { + "samplingParametersDeprecated": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 25, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "xhigh" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-opus-4-8", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-opus-4-8" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Opus 4.8，Anthropic Opus 系列最新旗舰", + "constraints": { + "samplingParametersDeprecated": true + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-opus-4-7", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-opus-4-7" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "Anthropic 当前最强通用模型，适合复杂推理和智能体编码任务", + "constraints": { + "samplingParametersDeprecated": true + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-opus", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-opus" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + } + }, + { + "modelRef": "anthropic:claude-sonnet-5-5", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-sonnet-5-5" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Sonnet 5.5，速度与智能平衡，默认 high effort，自适应思考", + "constraints": { + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + } + }, + "routing": { + "tier": "balanced", + "routingPriority": 18, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json", + "recordedFacts": { + "releasedAt": "2026-09-28" + } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-sonnet-5", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-sonnet-5" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Claude Sonnet 5，速度与智能平衡；思考默认开启，effort 支持到 max", + "constraints": { + "samplingParametersDeprecated": true + } + }, + "routing": { + "tier": "balanced", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "xhigh" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "anthropic:claude-sonnet-4-6", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-sonnet-4-6" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 64000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use" + ], + "serviceType": [ + "chat", + "computer_use" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Anthropic 高智能高速度模型，适合编码、工具使用和智能体任务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + } + }, + { + "modelRef": "anthropic:claude-sonnet-4-5", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-sonnet-4-5" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 64000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Anthropic 主力模型，长文本与代码任务表现优异" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + } + }, + { + "modelRef": "anthropic:claude-sonnet", + "identity": { + "vendor": "anthropic", + "canonicalModelId": "claude-sonnet" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/anthropic.json" + } + }, + { + "modelRef": "baichuan:Baichuan-M3-Plus", + "identity": { + "vendor": "baichuan", + "canonicalModelId": "Baichuan-M3-Plus" + }, + "facts": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "long_context", + "medical" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M3-Plus医疗增强模型，235B参数，幻觉率仅2.6%" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baichuan.json" + } + }, + { + "modelRef": "baichuan:Baichuan-M3", + "identity": { + "vendor": "baichuan", + "canonicalModelId": "Baichuan-M3" + }, + "facts": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "medical" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M3新一代开源医疗增强模型，235B参数" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baichuan.json" + } + }, + { + "modelRef": "baichuan:Baichuan-M2-Plus", + "identity": { + "vendor": "baichuan", + "canonicalModelId": "Baichuan-M2-Plus" + }, + "facts": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M2-Plus旗舰模型，最强推理和创作能力" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baichuan.json" + } + }, + { + "modelRef": "baichuan:Baichuan-M2", + "identity": { + "vendor": "baichuan", + "canonicalModelId": "Baichuan-M2" + }, + "facts": { + "contextWindow": 32000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "百川智能M2高性能通用模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baichuan.json" + } + }, + { + "modelRef": "baidu:ernie-5.0-thinking-latest", + "identity": { + "vendor": "baidu", + "canonicalModelId": "ernie-5.0-thinking-latest" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "vision", + "multimodal", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "百度最新旗舰模型，支持深度推理、多模态和工具调用" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baidu.json" + } + }, + { + "modelRef": "baidu:ernie-5.0", + "identity": { + "vendor": "baidu", + "canonicalModelId": "ernie-5.0" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "vision", + "multimodal", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.95, + "description": "百度 ERNIE 5.0 旗舰模型，128K 上下文，最大输出 65536 token" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baidu.json" + } + }, + { + "modelRef": "baidu:ernie-4.5-turbo-128k", + "identity": { + "vendor": "baidu", + "canonicalModelId": "ernie-4.5-turbo-128k" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 12288, + "capabilities": [ + "chat", + "code", + "vision", + "long_context", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.8, + "description": "百度高性价比长上下文模型，128K 窗口，最大输出 12288 token" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baidu.json" + } + }, + { + "modelRef": "baidu:ernie-4.5-turbo-20260402", + "identity": { + "vendor": "baidu", + "canonicalModelId": "ernie-4.5-turbo-20260402" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 12288, + "capabilities": [ + "chat", + "code", + "vision", + "long_context", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.8, + "description": "百度 ERNIE 4.5 Turbo 20260402 快照版，128K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baidu.json" + } + }, + { + "modelRef": "baidu:ernie-x1.1", + "identity": { + "vendor": "baidu", + "canonicalModelId": "ernie-x1.1" + }, + "facts": { + "contextWindow": 65536, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "math", + "code", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "百度 ERNIE X1.1 深度思考推理模型，64K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/baidu.json" + } + }, + { + "modelRef": "cohere:rerank-v4.0-fast", + "identity": { + "vendor": "cohere", + "canonicalModelId": "rerank-v4.0-fast" + }, + "facts": { + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Rerank 4 Fast" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:embed-v5.0-fast", + "identity": { + "vendor": "cohere", + "canonicalModelId": "embed-v5.0-fast" + }, + "facts": { + "capabilities": [ + "text_embedding", + "vision", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Embed 5 Fast", + "contextWindow": 128000, + "constraints": { + "dimensions": [ + 256, + 512, + 768, + 1024, + 1536, + 2048 + ], + "defaultDimension": 2048 + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json", + "recordedFacts": { + "releasedAt": "2026-09-30" + } + } + }, + { + "modelRef": "cohere:rerank-v4.0-pro", + "identity": { + "vendor": "cohere", + "canonicalModelId": "rerank-v4.0-pro" + }, + "facts": { + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Rerank 4 Pro" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:embed-v5.0-pro", + "identity": { + "vendor": "cohere", + "canonicalModelId": "embed-v5.0-pro" + }, + "facts": { + "capabilities": [ + "text_embedding", + "vision", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Embed 5 Pro", + "contextWindow": 128000, + "constraints": { + "dimensions": [ + 256, + 512, + 768, + 1024, + 1536, + 2048 + ], + "defaultDimension": 2048 + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json", + "recordedFacts": { + "releasedAt": "2026-09-30" + } + } + }, + { + "modelRef": "cohere:command-a-plus-05-2026", + "identity": { + "vendor": "cohere", + "canonicalModelId": "command-a-plus-05-2026" + }, + "facts": { + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "multilingual", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "Command A+", + "contextWindow": 128000, + "maxOutputTokens": 64000, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:north-mini-code-1-0", + "identity": { + "vendor": "cohere", + "canonicalModelId": "north-mini-code-1-0" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 64000, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "description": "Cohere North Mini Code 1.0，原厂API精确ID为north-mini-code-1-0；north-mini-code仅保留本仓历史配置匹配，不声明原厂接受该简称" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:north-small-translate-1-0", + "identity": { + "vendor": "cohere", + "canonicalModelId": "north-small-translate-1-0" + }, + "facts": { + "serviceType": [ + "chat" + ], + "capabilities": [ + "chat", + "multilingual" + ], + "description": "Cohere原厂机器翻译模型，支持50多种语言；官网窗口与输出为16k简写，本轮不猜测精确整数。只登记共享规格，原生接入与账号权限需独立验证" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:command-a-03-2025", + "identity": { + "vendor": "cohere", + "canonicalModelId": "command-a-03-2025" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 8000, + "capabilities": [ + "chat", + "code", + "tool_use", + "rag", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.3, + "description": "Cohere Command A 文本模型，256K 上下文，支持 RAG 和工具调用" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:command-r7b-12-2024", + "identity": { + "vendor": "cohere", + "canonicalModelId": "command-r7b-12-2024" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 4000, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "rag", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 0.3, + "description": "Cohere 小型高速对话模型，适合高吞吐、低延迟场景" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:embed-v4.0", + "identity": { + "vendor": "cohere", + "canonicalModelId": "embed-v4.0" + }, + "facts": { + "contextWindow": 128000, + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Cohere 最新 Embedding 模型，128K 上下文，多语言支持", + "constraints": { + "dimensions": [ + 256, + 512, + 1024, + 1536 + ], + "defaultDimension": 1536 + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "cohere:rerank-v3.5", + "identity": { + "vendor": "cohere", + "canonicalModelId": "rerank-v3.5" + }, + "facts": { + "contextWindow": 4096, + "capabilities": [ + "rerank", + "semantic_reranking" + ], + "serviceType": [ + "rerank" + ], + "description": "Cohere 语义重排序模型，用于检索结果精排" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/cohere.json" + } + }, + { + "modelRef": "deepseek:deepseek-v4-pro", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-v4-pro" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "DeepSeek V4 Pro，1.6T MoE 架构，100 万上下文" + }, + "routing": { + "tier": "flagship", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": true, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-v4-pro-0813", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-v4-pro-0813" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "DeepSeek V4 Pro 0813 版本，1.6T MoE 架构，100 万上下文" + }, + "routing": { + "tier": "flagship", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-chat", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-chat" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-reasoner", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-reasoner" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-flash", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-flash" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "DeepSeek V4.1 Flash，支持文本与图像输入，100 万上下文" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-v4-flash", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-v4-flash" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-v4-flash-0731", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-v4-flash-0731" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "DeepSeek V4 Flash 0731 固定版本，100 万上下文" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "high", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/deepseek.json" + } + }, + { + "modelRef": "deepseek:deepseek-v4-flash-vision-exp", + "identity": { + "vendor": "deepseek", + "canonicalModelId": "deepseek-v4-flash-vision-exp" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 384000, + "capabilities": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "DeepSeek V4 Flash 实验性视觉版本，支持文本与图像输入，文本能力与 V4 Flash 一致" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/deepseek.json", + "recordedFacts": { + "releasedAt": "2026-08-21" + } + } + }, + { + "modelRef": "google:gemini-embedding-2", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-embedding-2" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "vision", + "audio_understanding", + "video_understanding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "Gemini Embedding 2，原生多模态嵌入，8192 token 输入，共享文本/图像/音视频向量空间", + "constraints": { + "dimensions": [ + 768, + 1536, + 3072 + ], + "defaultDimension": 3072 + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.5-transcribe-live", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.5-transcribe-live" + }, + "facts": { + "capabilities": [ + "asr", + "realtime" + ], + "serviceType": [ + "asr" + ], + "description": "gemini-3.5-transcribe-live，官方模型目录；原生音频协议需客户端适配" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.5-transcribe", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.5-transcribe" + }, + "facts": { + "capabilities": [ + "asr" + ], + "serviceType": [ + "asr" + ], + "description": "gemini-3.5-transcribe，官方模型目录；原生音频协议需客户端适配" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.8-live-extended-thinking", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.8-live-extended-thinking" + }, + "facts": { + "capabilities": [ + "realtime", + "speech_to_speech", + "reasoning" + ], + "serviceType": [ + "realtime_voice" + ], + "description": "gemini-3.8-live-extended-thinking，官方模型目录；原生音频协议需客户端适配" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.8-live", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.8-live" + }, + "facts": { + "capabilities": [ + "realtime", + "speech_to_speech" + ], + "serviceType": [ + "realtime_voice" + ], + "description": "gemini-3.8-live，官方模型目录；原生音频协议需客户端适配" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.8-flash-lite-tts", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.8-flash-lite-tts" + }, + "facts": { + "capabilities": [ + "tts", + "fast" + ], + "serviceType": [ + "tts" + ], + "description": "gemini-3.8-flash-lite-tts，官方模型目录；原生音频协议需客户端适配" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.8-flash-tts", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.8-flash-tts" + }, + "facts": { + "capabilities": [ + "tts" + ], + "serviceType": [ + "tts" + ], + "description": "gemini-3.8-flash-tts，官方模型目录；原生音频协议需客户端适配", + "contextWindow": 8192, + "maxOutputTokens": 16384 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.8-flash", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.8-flash" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Gemini 3.8 Flash，稳定版，支持多模态输入和 1M 上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.7-flash", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.7-flash" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Gemini 3.7 Flash，稳定版，支持多模态输入和 1M 上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.5-flash-lite", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.5-flash-lite" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Gemini 3.5 Flash-Lite，稳定版，支持多模态输入和 1M 上下文" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 20, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.1-flash-lite-image", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.1-flash-lite-image" + }, + "facts": { + "contextWindow": 65536, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "image_generation", + "image_editing", + "tool_use" + ], + "serviceType": [ + "image_gen" + ], + "supportsReasoning": true, + "description": "Nano Banana 2 Lite，Google 面向高吞吐图像生成与编辑的轻量模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json", + "recordedFacts": { + "releasedAt": "2026-06-30" + } + } + }, + { + "modelRef": "google:gemini-3.5-flash", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.5-flash" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "audio_understanding", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Google Gemini 3.5 Flash，全模态高性价比模型，100 万上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.1-flash-image", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.1-flash-image" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 32768, + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "Google Nano Banana 2，最新文生图与编辑模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.1-pro-preview", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.1-pro-preview" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3-flash-preview", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3-flash-preview" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-2.5-pro", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-2.5-pro" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-2.5-flash", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-2.5-flash" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:gemini-3.1-flash-lite-preview", + "identity": { + "vendor": "google", + "canonicalModelId": "gemini-3.1-flash-lite-preview" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "Google Gemini 3.1 Flash-Lite Preview，低成本高吞吐多模态模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "google:text-embedding-005", + "identity": { + "vendor": "google", + "canonicalModelId": "text-embedding-005" + }, + "facts": { + "contextWindow": 2048, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "serviceType": [ + "embedding" + ], + "description": "Google 文本嵌入模型，768 维度" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/google.json" + } + }, + { + "modelRef": "happyhorse:happyhorse-1.1-t2v", + "identity": { + "vendor": "happyhorse", + "canonicalModelId": "happyhorse-1.1-t2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 文生视频，进一步提升文本词义理解、镜头调度与动态生成表现" + }, + "routing": { + "tier": "flagship", + "routingPriority": 116, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/happyhorse.json" + } + }, + { + "modelRef": "happyhorse:happyhorse-1.0-t2v", + "identity": { + "vendor": "happyhorse", + "canonicalModelId": "happyhorse-1.0-t2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.0 文生视频，具备高度还原的动态画面生成能力" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/happyhorse.json" + } + }, + { + "modelRef": "happyhorse:happyhorse-1.1-i2v", + "identity": { + "vendor": "happyhorse", + "canonicalModelId": "happyhorse-1.1-i2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 图生视频，进一步提升画面质感、动态表现与跨片段一致性" + }, + "routing": { + "tier": "flagship", + "routingPriority": 115, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/happyhorse.json" + } + }, + { + "modelRef": "happyhorse:happyhorse-1.0-i2v", + "identity": { + "vendor": "happyhorse", + "canonicalModelId": "happyhorse-1.0-i2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.0 图生视频，具备高度还原的动态画面生成能力" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/happyhorse.json" + } + }, + { + "modelRef": "happyhorse:happyhorse-1.1-r2v", + "identity": { + "vendor": "happyhorse", + "canonicalModelId": "happyhorse-1.1-r2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "reference_to_video", + "multi_image", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.1 参考生视频，支持最多9张图片参考，更加稳定的主体与场景参考" + }, + "routing": { + "tier": "flagship", + "routingPriority": 117, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/happyhorse.json" + } + }, + { + "modelRef": "happyhorse:happyhorse-1.0-r2v", + "identity": { + "vendor": "happyhorse", + "canonicalModelId": "happyhorse-1.0-r2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "reference_to_video", + "multi_image" + ], + "serviceType": [ + "video_gen" + ], + "description": "HappyHorse 1.0 参考生视频，支持最多9张图片参考" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/happyhorse.json" + } + }, + { + "modelRef": "kling:kling-3.0-turbo", + "identity": { + "vendor": "kling", + "canonicalModelId": "kling-3.0-turbo" + }, + "facts": { + "serviceType": [ + "video_gen" + ], + "capabilities": [ + "video_generation", + "image_to_video" + ], + "description": "可灵 3.0 Turbo，新版 /image-to-video/kling-3.0-turbo API，需独立协议适配" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/kling.json" + } + }, + { + "modelRef": "kling:kling-v3", + "identity": { + "vendor": "kling", + "canonicalModelId": "kling-v3" + }, + "facts": { + "serviceType": [ + "video_gen" + ], + "capabilities": [ + "video_generation", + "image_to_video" + ], + "description": "可灵 3.0，通过现有 /v1/videos/text2video 任务 API 的 model_name 调用" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/kling.json" + } + }, + { + "modelRef": "kling:kling-v2-5-turbo", + "identity": { + "vendor": "kling", + "canonicalModelId": "kling-v2-5-turbo" + }, + "facts": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "high_quality", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵最新视频生成模型，高性价比" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/kling.json" + } + }, + { + "modelRef": "kling:kling-v2-5-turbo-pro", + "identity": { + "vendor": "kling", + "canonicalModelId": "kling-v2-5-turbo-pro" + }, + "facts": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "ultra_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵高品质视频生成，支持 4K 分辨率" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/kling.json" + } + }, + { + "modelRef": "kling:kling-v2", + "identity": { + "vendor": "kling", + "canonicalModelId": "kling-v2" + }, + "facts": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵文生视频旗舰模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/kling.json" + } + }, + { + "modelRef": "kling:kling-v2-master", + "identity": { + "vendor": "kling", + "canonicalModelId": "kling-v2-master" + }, + "facts": { + "capabilities": [ + "video_generation", + "chinese_optimized", + "ultra_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "可灵最高品质视频生成，支持 4K 分辨率" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/kling.json" + } + }, + { + "modelRef": "minimax:MiniMax-H3-Max", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-H3-Max" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "MiniMax H3 Max，使用原生视频 V2 API（/v1/video_generation_v2），不适用 Anthropic Messages" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "maxVideoDuration": 15 + }, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-H3", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-H3" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video" + ], + "serviceType": [ + "video_gen" + ], + "description": "MiniMax H3，使用原生视频 V2 API（/v1/video_generation_v2），不适用 Anthropic Messages" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "maxVideoDuration": 15 + }, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M3.1-Flash-Preview", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M3.1-Flash-Preview" + }, + "facts": { + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "MiniMax M3.1 Flash Preview", + "contextWindow": 1000000, + "supportsReasoning": true, + "constraints": { + "thinkingOnly": true + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "minimax:MiniMax-M3", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M3" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 512000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "video_understanding", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "MiniMax M3 全模态基础模型，100 万上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2.7", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2.7" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2.5", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2.5" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "code", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2.7-highspeed", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2.7-highspeed" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "MiniMax M2.7 极速版，效果一致，输出更快，适合低延迟场景" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2.5-highspeed", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2.5-highspeed" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "MiniMax M2.5 高速推理版本（100TPS），适合低延迟场景" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2.1", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2.1" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "code", + "tool_use", + "reasoning" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "MiniMax 上一代主力模型，200K 上下文，性价比优秀" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-M2.1-highspeed", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-M2.1-highspeed" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "MiniMax M2.1 高速推理版本，适合低延迟场景" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:M2-her", + "identity": { + "vendor": "minimax", + "canonicalModelId": "M2-her" + }, + "facts": { + "contextWindow": 204800, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "roleplay" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "MiniMax 角色扮演对话模型，专为多轮角色对话设计" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-Text-01", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-Text-01" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.1, + "description": "MiniMax 百万级长上下文文本模型，擅长长文档处理" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:image-01", + "identity": { + "vendor": "minimax", + "canonicalModelId": "image-01" + }, + "facts": { + "capabilities": [ + "image_generation", + "subject_reference", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图/图生图模型，支持主题参考（角色一致性）" + }, + "routing": { + "tier": "balanced", + "routingPriority": 113, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:image-01-live", + "identity": { + "vendor": "minimax", + "canonicalModelId": "image-01-live" + }, + "facts": { + "capabilities": [ + "image_generation", + "style_control", + "subject_reference", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "MiniMax 文生图模型（Live 版），支持画风控制" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:speech-2.8-hd", + "identity": { + "vendor": "minimax", + "canonicalModelId": "speech-2.8-hd" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 最新高清语音合成，40+ 语言、100+ 预置音色" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:speech-2.8-turbo", + "identity": { + "vendor": "minimax", + "canonicalModelId": "speech-2.8-turbo" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成模型，速度优先，适合实时场景" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:speech-2.6-hd", + "identity": { + "vendor": "minimax", + "canonicalModelId": "speech-2.6-hd" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 高清语音合成（2.6 版），音质优秀，生成速度快" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:speech-2.6-turbo", + "identity": { + "vendor": "minimax", + "canonicalModelId": "speech-2.6-turbo" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "emotion_control", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成（2.6 版），低延迟响应" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:speech-02-hd", + "identity": { + "vendor": "minimax", + "canonicalModelId": "speech-02-hd" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "voice_clone", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 高清语音合成（历史版本），韵律稳定性优秀" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:speech-02-turbo", + "identity": { + "vendor": "minimax", + "canonicalModelId": "speech-02-turbo" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "MiniMax 快速语音合成（历史版本），增强小语种支持" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-Hailuo-2.3", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-Hailuo-2.3" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "camera_control", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "MiniMax 最新视频生成模型，支持 15 种运镜指令控制" + }, + "routing": { + "tier": "balanced", + "routingPriority": 102, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-Hailuo-2.3-Fast", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-Hailuo-2.3-Fast" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "camera_control", + "fast", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "海螺视频快速版本（仅图生视频），生成速度更快" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 101, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:T2V-01-Director", + "identity": { + "vendor": "minimax", + "canonicalModelId": "T2V-01-Director" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video", + "camera_control", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "导演级文生视频模型，精准运镜控制（历史版本）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:MiniMax-Hailuo-02", + "identity": { + "vendor": "minimax", + "canonicalModelId": "MiniMax-Hailuo-02" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "海螺视频 02，原生 1080P 生成（历史版本）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:S2V-01", + "identity": { + "vendor": "minimax", + "canonicalModelId": "S2V-01" + }, + "facts": { + "capabilities": [ + "video_generation", + "subject_reference", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "主体参考视频生成，基于参考图保持角色一致性" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:music-2.6", + "identity": { + "vendor": "minimax", + "canonicalModelId": "music-2.6" + }, + "facts": { + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成模型，支持音乐灵感描述和歌词输入；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:music-2.5+", + "identity": { + "vendor": "minimax", + "canonicalModelId": "music-2.5+" + }, + "facts": { + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成（2.5+ 版），音质提升；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:music-2.5", + "identity": { + "vendor": "minimax", + "canonicalModelId": "music-2.5" + }, + "facts": { + "capabilities": [ + "music_generation", + "lyrics_input", + "instrumental", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 音乐生成（2.5 版）；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "minimax:music-cover", + "identity": { + "vendor": "minimax", + "canonicalModelId": "music-cover" + }, + "facts": { + "capabilities": [ + "music_generation", + "cover", + "chinese_optimized" + ], + "serviceType": [ + "music_gen" + ], + "description": "MiniMax AI 翻唱模型，基于参考音频生成翻唱版本；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/minimax.json" + } + }, + { + "modelRef": "mistral:mistral-medium-3.5", + "identity": { + "vendor": "mistral", + "canonicalModelId": "mistral-medium-3.5" + }, + "facts": { + "contextWindow": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Mistral Medium 3.5，官方 API ID mistral-medium-3-5，256K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/mistral.json" + } + }, + { + "modelRef": "mistral:mistral-large-latest", + "identity": { + "vendor": "mistral", + "canonicalModelId": "mistral-large-latest" + }, + "facts": { + "contextWindow": 256000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/mistral.json" + } + }, + { + "modelRef": "mistral:mistral-small-latest", + "identity": { + "vendor": "mistral", + "canonicalModelId": "mistral-small-latest" + }, + "facts": { + "contextWindow": 262144, + "capabilities": [ + "chat", + "code", + "vision", + "reasoning", + "fast", + "tool_use" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 0.7 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/mistral.json" + } + }, + { + "modelRef": "mistral:codestral-latest", + "identity": { + "vendor": "mistral", + "canonicalModelId": "codestral-latest" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 32768, + "capabilities": [ + "code", + "fill_in_the_middle" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/mistral.json" + } + }, + { + "modelRef": "moonshot:kimi-k3", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "kimi-k3" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 1048576, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding" + ], + "serviceType": [ + "reasoning" + ], + "supportsReasoning": true, + "description": "Kimi K3，1M 上下文，始终开启思考，low/high/max 三档（默认 max）；输出上限与输入共享窗口，默认 max_completion_tokens=131072", + "constraints": { + "thinkingOnly": true, + "samplingParametersDeprecated": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:kimi-k2.7-code", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "kimi-k2.7-code" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Moonshot Kimi K2.7 Code 编程专用模型" + }, + "routing": { + "tier": "balanced", + "routingPriority": 50, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:kimi-k2.6", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "kimi-k2.6" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 45, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:kimi-k2.5", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "kimi-k2.5" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:kimi-k2-thinking", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "kimi-k2-thinking" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 16384, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:kimi-k2", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "kimi-k2" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:moonshot-v1-8k", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "moonshot-v1-8k" + }, + "facts": { + "contextWindow": 8192, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:moonshot-v1-32k", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "moonshot-v1-32k" + }, + "facts": { + "contextWindow": 32768, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "moonshot:moonshot-v1-128k", + "identity": { + "vendor": "moonshot", + "canonicalModelId": "moonshot-v1-128k" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "code", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/moonshot.json" + } + }, + { + "modelRef": "openai:gpt-6.1-sol", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-6.1-sol" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-6.1 Sol，1.05M 上下文，128K 最大输出；工具调用需要 Responses API，不支持关闭推理" + }, + "routing": { + "tier": "balanced", + "routingPriority": 4, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-image-2.5-sunburst", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-image-2.5-sunburst" + }, + "facts": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "图像生成与编辑模型，适合对编辑精度要求高的任务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-image-2.5-flare", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-image-2.5-flare" + }, + "facts": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering", + "vision", + "fast" + ], + "serviceType": [ + "image_gen" + ], + "description": "快速图像生成与编辑模型，适合日常高质量出图" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-6-sol", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-6-sol" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-6 Sol，1.05M 上下文，128K 最大输出" + }, + "routing": { + "tier": "balanced", + "routingPriority": 5, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-6-luna", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-6-luna" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-6 Luna，1.05M 上下文，128K 最大输出" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 5, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-6-astra", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-6-astra" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "constraints": { + "thinkingOnly": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-reserve", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-reserve" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "参数暂按 GPT-5.6 Luna 配置，待官方规格确认；保留手动选择，未经官方规格核验前不参与 Agent 自动选型" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 10, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-image-2", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-image-2" + }, + "facts": { + "contextWindow": 400000, + "capabilities": [ + "image_generation", + "image_editing", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "OpenAI GPT Image 2 文生图模型，支持高质量图片生成与编辑，400K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.5", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.5" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "flagship", + "routingPriority": 60, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.4", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.4" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 60, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.2", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.2" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.1", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.1" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5-mini" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "fast", + "tool_use", + "code", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5-nano", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5-nano" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "code", + "fast", + "tool_use" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4.1", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4.1" + }, + "facts": { + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4o", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4o" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4o-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4o-mini" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:o3", + "identity": { + "vendor": "openai", + "canonicalModelId": "o3" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "math", + "science", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:o3-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "o3-mini" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "math", + "code", + "fast" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:o4-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "o4-mini" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "math", + "code", + "tool_use", + "vision" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.2-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.2-pro" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "OpenAI GPT-5.2 Pro 专业版，最高性能" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5-pro" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 272000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "OpenAI GPT-5 Pro 专业版高性能推理模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4.1-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4.1-mini" + }, + "facts": { + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "OpenAI GPT-4.1 mini，1M 上下文，高性价比" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4.1-nano", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4.1-nano" + }, + "facts": { + "contextWindow": 1047576, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "code", + "ultra_long_context", + "fast", + "tool_use" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "description": "OpenAI GPT-4.1 nano，1M 上下文，极致性价比" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:text-embedding-3-small", + "identity": { + "vendor": "openai", + "canonicalModelId": "text-embedding-3-small" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "serviceType": [ + "embedding" + ], + "description": "OpenAI 轻量文本向量化模型，1536 维" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:text-embedding-3-large", + "identity": { + "vendor": "openai", + "canonicalModelId": "text-embedding-3-large" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search" + ], + "serviceType": [ + "embedding" + ], + "description": "OpenAI 高精度文本向量化模型，3072 维" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:tts-1", + "identity": { + "vendor": "openai", + "canonicalModelId": "tts-1" + }, + "facts": { + "capabilities": [ + "tts" + ], + "serviceType": [ + "tts" + ], + "description": "OpenAI 标准语音合成" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:tts-1-hd", + "identity": { + "vendor": "openai", + "canonicalModelId": "tts-1-hd" + }, + "facts": { + "capabilities": [ + "tts" + ], + "serviceType": [ + "tts" + ], + "description": "OpenAI 高清语音合成" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:whisper-1", + "identity": { + "vendor": "openai", + "canonicalModelId": "whisper-1" + }, + "facts": { + "capabilities": [ + "asr", + "multilingual" + ], + "serviceType": [ + "asr" + ], + "description": "OpenAI 通用语音识别" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:o3-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "o3-pro" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 100000, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "math", + "science" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "OpenAI o3-pro 高级推理模型，200K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:dall-e-3", + "identity": { + "vendor": "openai", + "canonicalModelId": "dall-e-3" + }, + "facts": { + "capabilities": [ + "image_generation", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "OpenAI 高质量文生图模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4o-realtime", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4o-realtime" + }, + "facts": { + "contextWindow": 32000, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "realtime" + ], + "serviceType": [ + "omni" + ], + "defaultTemperature": 0.7, + "description": "OpenAI 全能实时模型，支持语音 + 文本 + 视觉" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-4o-realtime-preview", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-4o-realtime-preview" + }, + "facts": { + "contextWindow": 32000, + "maxOutputTokens": 4096, + "capabilities": [ + "realtime", + "speech_to_speech", + "low_latency" + ], + "serviceType": [ + "realtime_voice" + ], + "defaultTemperature": 0.7, + "description": "OpenAI 实时语音交互模型（预览版）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-oss-120b", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-oss-120b" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "reasoning", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "description": "OpenAI 开源 120B 模型（通过 OpenRouter 提供，:free 层）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.6-sol", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.6-sol" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "flagship", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "low" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.6-sol-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.6-sol-pro" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-5.6 Sol 的 Pro 固定推理变体，由上游以 reasoning.mode=pro 提供" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json", + "recordedFacts": { + "releasedAt": "2026-07-09" + } + } + }, + { + "modelRef": "openai:gpt-5.6-terra", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.6-terra" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.6-terra-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.6-terra-pro" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-5.6 Terra 的 Pro 固定推理变体，由上游以 reasoning.mode=pro 提供" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json", + "recordedFacts": { + "releasedAt": "2026-07-09" + } + } + }, + { + "modelRef": "openai:gpt-5.6-luna", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.6-luna" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "lightweight", + "routingPriority": 10, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.6-luna-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.6-luna-pro" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "GPT-5.6 Luna 的 Pro 固定推理变体，由上游以 reasoning.mode=pro 提供" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json", + "recordedFacts": { + "releasedAt": "2026-07-09" + } + } + }, + { + "modelRef": "openai:gpt-5.4-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.4-mini" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "lightweight", + "routingPriority": 25, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "low" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.3-codex-spark", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.3-codex-spark" + }, + "facts": { + "contextWindow": 128000, + "capabilities": [ + "chat", + "code", + "tool_use", + "fast" + ], + "serviceType": [ + "chat" + ] + }, + "routing": { + "tier": "lightweight", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.4-nano", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.4-nano" + }, + "facts": { + "contextWindow": 400000, + "maxOutputTokens": 128000, + "description": "GPT-5.4 Nano；官网确认400000总窗口、128000最大输出。2026-10-01宣布deprecated，停止服务日2027-04-01；当前不据此退役。独立输入上限未由本页确认。" + }, + "routing": {}, + "lifecycle": { + "status": "deprecated", + "retiredAt": "2027-04-01", + "evidence": "https://developers.openai.com/api/docs/deprecations", + "checkedAt": "2026-10-03" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-realtime-2.1", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-realtime-2.1" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 32000, + "description": "官方精确型号与窗口/输出已核；独立实时协议适配与账号验收由接入面决定，不继承通用Chat合同。" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-realtime-2.1-mini", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-realtime-2.1-mini" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 32000, + "description": "官方精确型号与窗口/输出已核；独立实时协议适配与账号验收由接入面决定，不继承通用Chat合同。" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-realtime-translate", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-realtime-translate" + }, + "facts": { + "contextWindow": 16000, + "maxOutputTokens": 2000, + "description": "官方精确型号与窗口/输出已核；独立实时协议适配与账号验收由接入面决定，不继承通用Chat合同。" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "openai:gpt-5.4-pro", + "identity": { + "vendor": "openai", + "canonicalModelId": "gpt-5.4-pro" + }, + "facts": { + "contextWindow": 1050000, + "maxOutputTokens": 128000, + "description": "官方独立Pro规格；Responses协议限制由实际接入面执行，不沿用基础型号或其他代际family。" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/openai.json" + } + }, + { + "modelRef": "perplexity:sonar-pro", + "identity": { + "vendor": "perplexity", + "canonicalModelId": "sonar-pro" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "web_search", + "reasoning", + "citation" + ], + "serviceType": [ + "chat" + ], + "description": "Perplexity 旗舰搜索增强模型，内置实时联网搜索和引用，200K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/perplexity.json" + } + }, + { + "modelRef": "perplexity:sonar-reasoning-pro", + "identity": { + "vendor": "perplexity", + "canonicalModelId": "sonar-reasoning-pro" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "web_search", + "reasoning", + "deep_thinking" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "Perplexity 深度推理模型，内置联网搜索和深度思考" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/perplexity.json" + } + }, + { + "modelRef": "perplexity:sonar", + "identity": { + "vendor": "perplexity", + "canonicalModelId": "sonar" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 4096, + "capabilities": [ + "chat", + "web_search", + "citation", + "fast" + ], + "serviceType": [ + "chat" + ], + "description": "Perplexity 轻量搜索模型，低成本联网搜索" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/perplexity.json" + } + }, + { + "modelRef": "qwen:qwen3.7-max", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.7-max" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "Qwen3.7 Max，纯文本模型，100 万上下文，131072 最大输出；当前版本等同 2026-05-20 快照" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.8-max-prime", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.8-max-prime" + }, + "facts": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "Qwen3.8 Max Prime；OpenRouter 官方目录已提供该 ID，不推断国内 DashScope 同名接入" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-image-3.0", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-image-3.0" + }, + "facts": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-image-3.0-pro", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-image-3.0-pro" + }, + "facts": { + "capabilities": [ + "image_generation", + "image_editing", + "text_rendering" + ], + "serviceType": [ + "image_gen" + ], + "description": "qwen-image-3.0-pro" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.7-text-rerank", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.7-text-rerank" + }, + "facts": { + "capabilities": [ + "rerank", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "Qwen3.7 Text Rerank" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.7-text-embedding-flash", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.7-text-embedding-flash" + }, + "facts": { + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "qwen3.7-text-embedding-flash" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.7-text-embedding", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.7-text-embedding" + }, + "facts": { + "capabilities": [ + "text_embedding", + "multilingual" + ], + "serviceType": [ + "embedding" + ], + "description": "qwen3.7-text-embedding" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.8-omni-flash", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.8-omni-flash" + }, + "facts": { + "capabilities": [ + "chat", + "reasoning", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "Qwen3.8 Omni Flash", + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.8-max", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.8-max" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context", + "agent", + "vision", + "video_understanding" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "通义千问 Qwen3.8 Max 稳定版，100 万上下文，支持视觉与视频理解" + }, + "routing": { + "tier": "flagship", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "xhigh" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.8-flash", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.8-flash" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "multilingual", + "tool_use", + "long_context", + "agent", + "vision", + "video_understanding", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "通义千问 Qwen3.8 Flash，高性价比多模态推理模型，支持 100 万上下文及文本、图像、视频输入" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultMode": "xhigh" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true, + "thinkingMaxTokens": 262144 + }, + "sourceFile": "compute/model-specs/qwen.json", + "recordedFacts": { + "releasedAt": "2026-08-26" + } + } + }, + { + "modelRef": "qwen:qwen3.8-max-preview", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.8-max-preview" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context", + "agent", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "通义千问 Qwen3.8 Max 预览版，100 万上下文，支持视觉理解" + }, + "routing": { + "tier": "flagship", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.7-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.7-plus" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "通义千问 Qwen3.7 Plus 高性价比多模态模型，100 万上下文" + }, + "routing": { + "tier": "balanced", + "routingPriority": 35, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-max", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-max" + }, + "facts": { + "contextWindow": 32768, + "maxOutputTokens": 8192, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-plus" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "thinking_mode", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-turbo", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-turbo" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "long_context", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-max", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-max" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.5-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.5-plus" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.6-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.6-plus" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.6-flash", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.6-flash" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "fast" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 0.6, + "supportsReasoning": true, + "description": "通义千问3.6 Flash，高性价比多模态模型，100万上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-long", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-long" + }, + "facts": { + "contextWindow": 10000000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "ultra_long_context", + "document_processing" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "通义千问 Long，支持 1000 万上下文，适合超长文档处理" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-vl-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-vl-plus" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "ocr", + "chart_analysis", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 0.7, + "description": "通义千问3代视觉模型Plus版，支持图像理解" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-vl-flash", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-vl-flash" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "image_understanding", + "fast", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 0.7, + "description": "通义千问3代视觉模型Flash版，高性价比" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:text-embedding-v3", + "identity": { + "vendor": "qwen", + "canonicalModelId": "text-embedding-v3" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "chinese_optimized" + ], + "serviceType": [ + "embedding" + ], + "description": "阿里云文本向量化模型 v3" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:text-embedding-v4", + "identity": { + "vendor": "qwen", + "canonicalModelId": "text-embedding-v4" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag" + ], + "serviceType": [ + "embedding" + ], + "description": "阿里云百炼文本向量模型v4，支持多维度与多语种" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-rerank", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-rerank" + }, + "facts": { + "contextWindow": 120000, + "capabilities": [ + "rerank", + "semantic_reranking", + "multilingual" + ], + "serviceType": [ + "rerank" + ], + "description": "检索重排序，支持 100+ 语种" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:cosyvoice-v2", + "identity": { + "vendor": "qwen", + "canonicalModelId": "cosyvoice-v2" + }, + "facts": { + "capabilities": [ + "tts", + "multilingual", + "chinese_optimized" + ], + "serviceType": [ + "tts" + ], + "description": "通义语音合成，支持中英文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:paraformer-v2", + "identity": { + "vendor": "qwen", + "canonicalModelId": "paraformer-v2" + }, + "facts": { + "capabilities": [ + "asr", + "chinese_optimized" + ], + "serviceType": [ + "asr" + ], + "description": "通义语音识别" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:wan2.7-image-pro", + "identity": { + "vendor": "qwen", + "canonicalModelId": "wan2.7-image-pro" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相最新旗舰文生图模型，支持 4K 分辨率" + }, + "routing": { + "tier": "flagship", + "routingPriority": 100, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:wan2.7-image", + "identity": { + "vendor": "qwen", + "canonicalModelId": "wan2.7-image" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相最新标准文生图模型，高画质" + }, + "routing": { + "tier": "balanced", + "routingPriority": 107, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:wan2.6-t2i", + "identity": { + "vendor": "qwen", + "canonicalModelId": "wan2.6-t2i" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相 2.6 文生图专用模型，稳定可靠" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:wan2.2-t2i-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "wan2.2-t2i-plus" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相 2.2 增强版，画质与速度平衡" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:wan2.2-t2i-flash", + "identity": { + "vendor": "qwen", + "canonicalModelId": "wan2.2-t2i-flash" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized", + "fast" + ], + "serviceType": [ + "image_gen" + ], + "description": "万相 2.2 快速版，生成速度极快" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:wan2.6-t2v", + "identity": { + "vendor": "qwen", + "canonicalModelId": "wan2.6-t2v" + }, + "facts": { + "capabilities": [ + "video_generation", + "chinese_optimized" + ], + "serviceType": [ + "video_gen" + ], + "description": "万相 2.6 文生视频模型，支持文生视频和图生视频" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:cosyvoice-clone", + "identity": { + "vendor": "qwen", + "canonicalModelId": "cosyvoice-clone" + }, + "facts": { + "capabilities": [ + "voice_clone", + "chinese_optimized", + "few_shot" + ], + "serviceType": [ + "voice_clone" + ], + "description": "通义声音复刻，少量样本即可克隆" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-omni-turbo", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-omni-turbo" + }, + "facts": { + "contextWindow": 32768, + "maxOutputTokens": 2048, + "capabilities": [ + "chat", + "vision", + "tts", + "asr", + "multimodal", + "tool_use" + ], + "serviceType": [ + "omni" + ], + "defaultTemperature": 0.7, + "description": "通义全能模型，支持音视频文本多模态" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-max-trans", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-max-trans" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 8192, + "capabilities": [ + "translation", + "multilingual", + "chinese_optimized" + ], + "serviceType": [ + "translation" + ], + "defaultTemperature": 0.65, + "description": "通义千问翻译，中英日韩等多语种互译" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-coder", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-coder" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "code", + "long_context", + "agent", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "通义千问3 Coder 编程模型（通过 OpenRouter 提供，:free 层）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.5-35b-a3b", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.5-35b-a3b" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3.5 35B-A3B MoE 模型（35B 总参/3B 激活），131K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3.5-27b", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3.5-27b" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3.5 27B Dense 模型，131K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-coder-480b", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-coder-480b" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "code", + "long_context", + "agent", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3-Coder-480B-A35B，480B MoE（35B 激活），256K 上下文，面向 Agentic Coding" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-235b", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-235b" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "code", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 0.7, + "description": "Qwen3-235B-A22B-Instruct-2507，235B MoE 通用指令模型，262K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-image-2.0-pro", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-image-2.0-pro" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized", + "high_quality" + ], + "serviceType": [ + "image_gen" + ], + "description": "千问图像生成 2.0 Pro 旗舰版，高画质多风格" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen-image-2.0", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen-image-2.0" + }, + "facts": { + "capabilities": [ + "image_generation", + "chinese_optimized" + ], + "serviceType": [ + "image_gen" + ], + "description": "千问图像生成 2.0 标准版" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-coder-plus", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-coder-plus" + }, + "facts": { + "description": "精确绑定已有接入型号，参数未核实；identity-only规格不继承旧family能力、窗口、协议或自动路由资格。" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "qwen:qwen3-coder-next", + "identity": { + "vendor": "qwen", + "canonicalModelId": "qwen3-coder-next" + }, + "facts": { + "description": "精确绑定已有接入型号，参数未核实；identity-only规格不继承旧family能力、窗口、协议或自动路由资格。" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/qwen.json" + } + }, + { + "modelRef": "stability:stable-audio-3.0", + "identity": { + "vendor": "stability", + "canonicalModelId": "stable-audio-3.0" + }, + "facts": { + "capabilities": [ + "music_generation", + "audio_generation" + ], + "serviceType": [ + "music_gen" + ], + "description": "Stable Audio 3.0，通过原生 /v2beta/audio/stable-audio API 提供，不使用 Stability 图像协议" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/stability.json", + "recordedFacts": { + "releasedAt": "2026-05-20" + } + } + }, + { + "modelRef": "stability:stable-diffusion-3.5-large", + "identity": { + "vendor": "stability", + "canonicalModelId": "stable-diffusion-3.5-large" + }, + "facts": { + "capabilities": [ + "image_generation", + "high_quality", + "controllable" + ], + "serviceType": [ + "image_gen" + ], + "description": "Stability AI 旗舰文生图模型，支持多种分辨率和风格" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/stability.json" + } + }, + { + "modelRef": "stealth:ox-alpha", + "identity": { + "vendor": "stealth", + "canonicalModelId": "ox-alpha" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "image_understanding", + "video_understanding" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "Ox Alpha 历史匿名模型兼容记录；当前官方目录未列出。旧记录曾关联 GLM-5.3-Flash，但本轮未取得官方身份揭示证据，不下发未证实的原厂身份映射；历史参数不代表当前可用性", + "constraints": { + "thinkingOnly": true + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/stealth.json", + "recordedFacts": { + "releasedAt": "2026-08-21" + } + } + }, + { + "modelRef": "tencent:hy4-preview", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hy4-preview" + }, + "facts": { + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "description": "Tencent Hy4 Preview，通过 TokenHub 提供；既有 Hunyuan Provider 端点可用性尚未确认", + "contextWindow": 1048576 + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json", + "recordedFacts": { + "releasedAt": "2026-08-28" + } + } + }, + { + "modelRef": "tencent:hy3", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hy3" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 0.9, + "supportsReasoning": true, + "description": "腾讯混元 Hy3 正式版，295B MoE（21B active），支持可配置推理强度" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json", + "recordedFacts": { + "releasedAt": "2026-07-06" + } + } + }, + { + "modelRef": "tencent:hy3-preview", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hy3-preview" + }, + "facts": { + "contextWindow": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "腾讯混元 Hy3 预览版，高效 MoE 模型" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "tencent:hunyuan-2.0-thinking-20251109", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hunyuan-2.0-thinking-20251109" + }, + "facts": { + "contextWindow": 196608, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "腾讯混元最新推理模型，MoE架构406B总参数，深度思考" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "tencent:hunyuan-2.0-instruct-20251111", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hunyuan-2.0-instruct-20251111" + }, + "facts": { + "contextWindow": 147456, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "code", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "腾讯混元 2.0 指令模型，支持联网搜索和 Function Calling" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "tencent:hunyuan-turbos-latest", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hunyuan-turbos-latest" + }, + "facts": { + "contextWindow": 32768, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "fast", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "腾讯混元高速模型，支持推理和工具调用" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "tencent:hunyuan-t1-latest", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hunyuan-t1-latest" + }, + "facts": { + "contextWindow": 262144, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "腾讯混元 T1 深度推理模型，256K 上下文，专攻数学/代码/复杂推理" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "tencent:hunyuan-t1-vision", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hunyuan-t1-vision" + }, + "facts": { + "contextWindow": 131072, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "vision", + "image_understanding" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "腾讯混元 T1 视觉推理模型，支持图片理解 + 深度推理" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "tencent:hunyuan-turbos-vision", + "identity": { + "vendor": "tencent", + "canonicalModelId": "hunyuan-turbos-vision" + }, + "facts": { + "contextWindow": 32768, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "image_understanding", + "fast", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "腾讯混元 Turbo S 视觉模型，支持图片理解，高速推理" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/tencent.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-evolving", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-evolving" + }, + "facts": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "Seed 系列持续迭代模型，通过固定 ID 获取最新版本，2026-09-09 更新" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-2.1-pro", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-2.1-pro" + }, + "facts": { + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "description": "火山方舟豆包 Seed 2.1 Pro，生产级 Coding、Agent 与多模态旗舰模型", + "contextWindow": 256000, + "maxOutputTokens": 262144, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "long_context", + "tool_use", + "agent", + "deep_thinking" + ], + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-2.0-pro", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-2.0-pro" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 Seed 2.0 旗舰模型，256K 上下文，支持复杂推理、Agent 和多模态" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-2.0-lite", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-2.0-lite" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "audio_understanding", + "video_understanding" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "豆包 Seed 2.0 全模态均衡模型，支持视频/图像/音频/文本统一理解" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-2.0-mini", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-2.0-mini" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision", + "audio_understanding", + "video_understanding", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "豆包 Seed 2.0 全模态轻量模型，低延迟高并发，支持全模态理解" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-2.0-code", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-2.0-code" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "long_context", + "tool_use", + "vision", + "deep_thinking" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "豆包 Seed 2.0 编程模型，面向 Agentic Coding 场景优化" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-1.8", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-1.8" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.8 旗舰模型，256K 上下文，支持深度思考和多模态" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-1.6", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-1.6" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 旗舰模型，256K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-1.6-thinking", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-1.6-thinking" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "火山引擎豆包 1.6 深度思考模型，256K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-1.6-flash", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-1.6-flash" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 16000, + "capabilities": [ + "chat", + "code", + "fast", + "long_context", + "vision" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 快速版，成本较前代下降 63%" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-1.6-lite", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-1.6-lite" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "code", + "fast" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 轻量版，极致性价比" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-1.6-vision", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-1.6-vision" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "gui_agent", + "long_context" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包 1.6 视觉模型，256K 上下文，支持视频理解" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-code", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-code" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 32768, + "capabilities": [ + "code", + "reasoning", + "long_context", + "deep_thinking" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎豆包编程专用模型，256K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:deepseek-v3.2", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "deepseek-v3.2" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 32000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "火山引擎托管的 DeepSeek V3.2，128K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:deepseek-r1", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "deepseek-r1" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 65536, + "capabilities": [ + "reasoning", + "math", + "code", + "science", + "deep_thinking" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "火山引擎托管的 DeepSeek R1 推理模型，128K 上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-embedding", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-embedding" + }, + "facts": { + "contextWindow": 4096, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "chinese_optimized" + ], + "serviceType": [ + "embedding" + ], + "description": "火山引擎豆包标准向量模型，默认 2048 维度，可选 512/1024/2048" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-embedding-large", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-embedding-large" + }, + "facts": { + "contextWindow": 4096, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "chinese_optimized" + ], + "serviceType": [ + "embedding" + ], + "description": "火山引擎豆包大号向量模型，默认 2048 维度，可选 512/1024/2048/4096" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:volc-mega-tts-clone", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "volc-mega-tts-clone" + }, + "facts": { + "capabilities": [ + "voice_clone", + "chinese_optimized", + "few_shot" + ], + "serviceType": [ + "voice_clone" + ], + "description": "火山引擎声音复刻，支持极少样本克隆" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:volc-realtime-voice", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "volc-realtime-voice" + }, + "facts": { + "capabilities": [ + "realtime", + "speech_to_speech", + "low_latency", + "chinese_optimized" + ], + "serviceType": [ + "realtime_voice" + ], + "description": "火山引擎实时语音交互" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:volc-simultaneous", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "volc-simultaneous" + }, + "facts": { + "capabilities": [ + "simultaneous_interpretation", + "chinese_optimized", + "low_latency" + ], + "serviceType": [ + "simultaneous_interpret" + ], + "description": "火山引擎同声传译，支持中英双向" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:volc-translation", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "volc-translation" + }, + "facts": { + "capabilities": [ + "translation", + "multilingual", + "chinese_optimized" + ], + "serviceType": [ + "translation" + ], + "description": "火山引擎机器翻译，支持多语种互译" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seedance-2.0", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seedance-2.0" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "video_editing", + "video_extension", + "audio_generation", + "first_last_frame" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 2.0 豆包最强视频生成模型，4-15秒 2K 视频，原生同步音频，支持最多9张图片+3个视频+3个音频参考输入，~$0.14/秒" + }, + "routing": { + "tier": "flagship", + "routingPriority": 120, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seedance-2.0-fast", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seedance-2.0-fast" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "video_editing", + "video_extension", + "audio_generation", + "first_last_frame", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 2.0 快速版，音画同生，多模态生视频+编辑+延长" + }, + "routing": { + "tier": "balanced", + "routingPriority": 119, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seedance-2.0-mini", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seedance-2.0-mini" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "video_editing", + "video_extension", + "audio_generation", + "first_last_frame", + "fast", + "cost_effective" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 2.0 轻量版，音画同生，高性价比" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 118, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seedance-1.5-pro", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seedance-1.5-pro" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "first_last_frame", + "audio_generation" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 1.5 Pro 音画同生，12秒视频，多参考图输入（即将下线）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seedance-1.0-pro", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seedance-1.0-pro" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "first_last_frame" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 1.0 Pro 文生视频/图生视频，支持首尾帧控制" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seedance-1.0-pro-fast", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seedance-1.0-pro-fast" + }, + "facts": { + "capabilities": [ + "video_generation", + "image_to_video", + "text_to_video", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "Seedance 1.0 Pro 快速版，图生视频-首帧，5秒 720P 仅需 10 秒" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "volcengine:doubao-seed-2.1-turbo", + "identity": { + "vendor": "volcengine", + "canonicalModelId": "doubao-seed-2.1-turbo" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "vision" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/volcengine.json" + } + }, + { + "modelRef": "wan:wan3.0-video-prime", + "identity": { + "vendor": "wan", + "canonicalModelId": "wan3.0-video-prime" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "high_quality", + "fast" + ], + "serviceType": [ + "video_gen" + ], + "description": "万相 3.0 全能参考视频生成高速版，能力对齐标准版并显著提升端到端生成速度" + }, + "routing": { + "tier": "flagship", + "routingPriority": 109, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "maxVideoDuration": 30, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ] + }, + "sourceFile": "compute/model-specs/wan.json", + "recordedFacts": { + "releasedAt": "2026-08-06" + } + } + }, + { + "modelRef": "wan:wan3.0-video", + "identity": { + "vendor": "wan", + "canonicalModelId": "wan3.0-video" + }, + "facts": { + "capabilities": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "high_quality" + ], + "serviceType": [ + "video_gen" + ], + "description": "万相 3.0 全能参考视频生成标准版，统一支持文生、首帧/首尾帧图生及多类型参考生视频" + }, + "routing": { + "tier": "flagship", + "routingPriority": 110, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "maxVideoDuration": 30, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ] + }, + "sourceFile": "compute/model-specs/wan.json", + "recordedFacts": { + "releasedAt": "2026-08-06" + } + } + }, + { + "modelRef": "xai:grok-4.7", + "identity": { + "vendor": "xai", + "canonicalModelId": "grok-4.7" + }, + "facts": { + "contextWindow": 500000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "xAI Grok 4.7，500K 上下文，文本与图像输入，推理努力等级 low 到 xhigh" + }, + "routing": { + "tier": "flagship", + "routingPriority": 30, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xai.json" + } + }, + { + "modelRef": "xai:grok-4.5", + "identity": { + "vendor": "xai", + "canonicalModelId": "grok-4.5" + }, + "facts": { + "contextWindow": 500000, + "maxOutputTokens": 450000, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "supportsReasoning": true, + "description": "xAI Grok 4.5，面向编程、知识工作和 STEM 的旗舰推理模型", + "constraints": { + "thinkingOnly": true + } + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/xai.json", + "recordedFacts": { + "releasedAt": "2026-07-08" + } + } + }, + { + "modelRef": "xai:grok-4-3", + "identity": { + "vendor": "xai", + "canonicalModelId": "grok-4-3" + }, + "facts": { + "contextWindow": 1000000, + "capabilities": [ + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "xAI Grok 4.3 推理模型，100 万上下文" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xai.json" + } + }, + { + "modelRef": "xai:grok-4-1-fast-reasoning", + "identity": { + "vendor": "xai", + "canonicalModelId": "grok-4-1-fast-reasoning" + }, + "facts": { + "contextWindow": 2000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "fast", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xai.json" + } + }, + { + "modelRef": "xai:grok-4.20-0309-reasoning", + "identity": { + "vendor": "xai", + "canonicalModelId": "grok-4.20-0309-reasoning" + }, + "facts": { + "contextWindow": 2000000, + "maxOutputTokens": 16384, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xai.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2.6-pro", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.6-pro" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "小米全模态旗舰推理模型，适合复杂项目与长程智能体任务" + }, + "routing": { + "tier": "flagship", + "routingPriority": 28, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "releasedAt": "2026-09-22" + } + } + }, + { + "modelRef": "xiaomi:mimo-v2.6-flash", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.6-flash" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "小米全模态高效推理模型，适合高频调用与规模化任务" + }, + "routing": { + "tier": "balanced", + "routingPriority": 28, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "releasedAt": "2026-09-22" + } + } + }, + { + "modelRef": "xiaomi:mimo-v2.6-pro-ultraspeed", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.6-pro-ultraspeed" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "code", + "vision", + "audio_understanding", + "video_understanding", + "tool_use", + "agent", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "MiMo-V2.6-Pro 的超高速服务，需向小米申请定制配额" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "releasedAt": "2026-09-22" + } + } + }, + { + "modelRef": "xiaomi:mimo-x-flash-preview", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-x-flash-preview" + }, + "facts": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "小米 MiMo-X Flash 内测模型，1M 上下文，支持图片识别" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-x-pro-preview", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-x-pro-preview" + }, + "facts": { + "contextWindow": 1000000, + "capabilities": [ + "chat", + "vision" + ], + "serviceType": [ + "chat", + "vision" + ], + "description": "小米 MiMo-X Pro 内测模型，1M 上下文，支持图片识别" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2.5-pro", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-pro" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo 旗舰推理模型，支持文本/工具调用" + }, + "routing": { + "tier": "flagship", + "routingPriority": 50, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2.5", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo 全模态模型，支持图像/音频/视频理解" + }, + "routing": { + "tier": "lightweight", + "routingPriority": 40, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "low" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2-pro", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2-pro" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "reasoning" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo V2 Pro（自 2026-06-01 转发至 V2.5 Pro）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "retiredAt": "2026-06-30" + } + } + }, + { + "modelRef": "xiaomi:mimo-v2-omni", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2-omni" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "vision", + "tool_use", + "code" + ], + "serviceType": [ + "chat", + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo V2 全模态（自 2026-06-01 转发至 V2.5）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "retiredAt": "2026-06-30" + } + } + }, + { + "modelRef": "xiaomi:mimo-v2-flash", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2-flash" + }, + "facts": { + "contextWindow": 256000, + "maxOutputTokens": 65536, + "capabilities": [ + "chat", + "reasoning", + "tool_use", + "code", + "fast" + ], + "serviceType": [ + "fast" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "小米 MiMo 快速版，适合高频低延迟任务" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2.5-tts", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-tts" + }, + "facts": { + "contextWindow": 8192, + "maxOutputTokens": 8192, + "capabilities": [ + "tts", + "multilingual", + "style_control" + ], + "serviceType": [ + "tts" + ], + "description": "小米 MiMo 语音合成，支持多语言与风格控制", + "constraints": { + "modelOrigin": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-tts", + "family": "mimo-tts" + }, + "speech": { + "protocol": "xiaomi-mimo-chat-tts", + "protocolVersion": "1", + "modes": [ + "batch", + "streaming" + ], + "outputFormats": [ + "wav", + "mp3", + "pcm16le" + ], + "sampleRates": [ + 24000 + ], + "defaultVoice": "mimo_default", + "voices": [ + { + "id": "mimo_default", + "label": "MiMo 默认" + }, + { + "id": "冰糖", + "label": "冰糖" + }, + { + "id": "茉莉", + "label": "茉莉" + }, + { + "id": "苏打", + "label": "苏打" + }, + { + "id": "白桦", + "label": "白桦" + }, + { + "id": "Mia", + "label": "Mia" + }, + { + "id": "Chloe", + "label": "Chloe" + }, + { + "id": "Milo", + "label": "Milo" + }, + { + "id": "Dean", + "label": "Dean" + } + ] + } + } + }, + "routing": { + "tier": "lightweight", + "routingPriority": 114, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "releasedAt": "2026-04-23" + } + } + }, + { + "modelRef": "xiaomi:mimo-v2.5-asr", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-asr" + }, + "facts": { + "contextWindow": 8192, + "maxOutputTokens": 2048, + "capabilities": [ + "asr", + "multilingual" + ], + "serviceType": [ + "asr" + ], + "description": "小米 MiMo 语音识别模型，支持中英文语言提示与整段音频识别", + "constraints": { + "modelOrigin": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-asr", + "family": "mimo-asr" + }, + "speech": { + "protocol": "xiaomi-mimo-chat-asr", + "protocolVersion": "1", + "modes": [ + "batch" + ], + "inputFormats": [ + "wav", + "mp3" + ], + "languageHints": true + } + } + }, + "routing": { + "tier": "lightweight", + "routingPriority": 123, + "eligibleForAgent": false, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto" + ], + "defaultMode": "auto" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json", + "recordedFacts": { + "releasedAt": "2026-06-02" + } + } + }, + { + "modelRef": "xiaomi:mimo-v2.5-tts-voiceclone", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-tts-voiceclone" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "tts", + "voice_clone", + "multilingual" + ], + "serviceType": [ + "tts", + "voice_clone" + ], + "description": "小米 MiMo 声音复刻合成" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2.5-tts-voicedesign", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2.5-tts-voicedesign" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "tts", + "voice_design", + "multilingual" + ], + "serviceType": [ + "tts" + ], + "description": "小米 MiMo 自定义音色设计合成" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xiaomi:mimo-v2-tts", + "identity": { + "vendor": "xiaomi", + "canonicalModelId": "mimo-v2-tts" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "tts", + "multilingual" + ], + "serviceType": [ + "tts" + ], + "description": "小米 MiMo V2 语音合成（旧版）" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xiaomi.json" + } + }, + { + "modelRef": "xunfei:spark-x", + "identity": { + "vendor": "xunfei", + "canonicalModelId": "spark-x" + }, + "facts": { + "contextWindow": 65536, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "math", + "code", + "tool_use" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": 1.2, + "supportsReasoning": true, + "description": "讯飞星火 X2，/x2 接入合同：输入 65536 / 输出 131072；spark-x 同名在 /v2 上指向 X1.5，不能复用本输出限制" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xunfei.json" + } + }, + { + "modelRef": "xunfei:4.0Ultra", + "identity": { + "vendor": "xunfei", + "canonicalModelId": "4.0Ultra" + }, + "facts": { + "contextWindow": 32768, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "description": "讯飞星火最强通用模型，支持推理、代码和工具调用" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/xunfei.json" + } + }, + { + "modelRef": "zhipu:glm-5.3-flashx", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5.3-flashx" + }, + "facts": { + "contextWindow": 1000000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "vision", + "video_understanding", + "tool_use", + "long_context", + "fast" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "supportsReasoning": true, + "description": "Z.ai GLM-5.3-FlashX，高速多模态推理模型，100 万上下文，始终开启推理", + "constraints": { + "thinkingOnly": true + } + }, + "routing": { + "tier": "balanced", + "routingPriority": 49, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "zhipu:glm-5.3", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5.3" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "math", + "multilingual", + "tool_use", + "agent", + "long_context" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "constraints": { + "thinkingOnly": true + } + }, + "routing": { + "tier": "flagship", + "routingPriority": 50, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/zhipu.json", + "recordedFacts": { + "releasedAt": "2026-08-16" + } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "zhipu:glm-5.3-flash", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5.3-flash" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 131072, + "capabilities": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "image_understanding", + "video_understanding" + ], + "serviceType": [ + "chat", + "reasoning", + "vision" + ], + "defaultTemperature": 1, + "defaultTopP": 0.95, + "supportsReasoning": true, + "description": "智谱 GLM-5.3-Flash 原生多模态模型，混合稀疏与线性注意力，面向高效编码与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "constraints": { + "thinkingOnly": true + } + }, + "routing": { + "tier": "balanced", + "routingPriority": 48, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "low", + "high", + "max" + ], + "defaultMode": "max" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": { + "thinkingDefault": true + }, + "sourceFile": "compute/model-specs/zhipu.json", + "recordedFacts": { + "releasedAt": "2026-08-26" + } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] + }, + { + "modelRef": "zhipu:glm-5.2", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5.2" + }, + "facts": { + "contextWindow": 1048576, + "maxOutputTokens": 32768, + "capabilities": [ + "reasoning", + "deep_thinking", + "code", + "multilingual", + "tool_use", + "long_context" + ], + "serviceType": [ + "chat" + ], + "supportsReasoning": true, + "description": "智谱 GLM-5.2 大推理模型，100 万上下文" + }, + "routing": { + "tier": "flagship", + "routingPriority": 45, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "high" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-5.1", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5.1" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": { + "tier": "balanced", + "routingPriority": 65, + "eligibleForAgent": true, + "defaultReference": false, + "reasoning": { + "supportedModes": [ + "auto", + "off", + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultMode": "medium" + } + }, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-5", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-5-turbo", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5-turbo" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-4.7", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-4.7" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-4.6", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-4.6" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "tool_use" + ], + "serviceType": [ + "chat" + ], + "defaultTemperature": 1, + "supportsReasoning": true + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-4.7-thinking", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-4.7-thinking" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "reasoning", + "math", + "code", + "deep_thinking", + "long_context" + ], + "serviceType": [ + "reasoning" + ], + "defaultTemperature": null, + "supportsReasoning": true, + "description": "智谱GLM-4.7深度思考模式，交错式/保留式/轮级思考" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-5v-turbo", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-5v-turbo" + }, + "facts": { + "contextWindow": 200000, + "maxOutputTokens": 128000, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "file_understanding", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1, + "supportsReasoning": true, + "description": "智谱首个多模态 Coding 基座模型，支持视频、图像、文本和文件输入" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:glm-4.6v", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "glm-4.6v" + }, + "facts": { + "contextWindow": 128000, + "maxOutputTokens": 32768, + "capabilities": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "long_context", + "tool_use" + ], + "serviceType": [ + "vision" + ], + "defaultTemperature": 1, + "description": "智谱GLM-4.6V多模态版，106B/12B MoE，支持图像视频理解" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + }, + { + "modelRef": "zhipu:embedding-3", + "identity": { + "vendor": "zhipu", + "canonicalModelId": "embedding-3" + }, + "facts": { + "contextWindow": 8192, + "capabilities": [ + "text_embedding", + "semantic_search", + "rag", + "custom_dimensions" + ], + "serviceType": [ + "embedding" + ], + "description": "智谱嵌入模型v3，支持自定义维度；单条输入最多 3072 tokens" + }, + "routing": {}, + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "recordedExtra": {}, + "sourceFile": "compute/model-specs/zhipu.json" + } + } + ], + "accesses": [ + { + "id": "provider-anthropic-claude-plan-001", + "category": "providers", + "connection": { + "id": "provider-anthropic-claude-plan-001", + "provider": "anthropic-claude", + "brandGroup": "anthropic", + "label": "Claude 订阅 (Claude Code)", + "baseUrl": "https://api.anthropic.com", + "apiFormat": "anthropic-messages", + "apiKeyRef": "anthropic-claude-oauth-token", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "accessMode": "coding-plan", + "credentialSource": "claude-oauth", + "requiredClientVersion": "10.0.83", + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://claude.ai/settings/usage" + } + }, + "services": [ + "chat", + "reasoning" + ], + "tombstones": [ + "claude-opus-4-8", + "claude-fable-5", + "claude-haiku-4-5", + "claude-haiku-4-5-20251001" + ] + }, + "models": [ + { + "apiModelId": "claude-opus-5-5", + "modelRef": "anthropic:claude-opus-5-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic-claude.json#claude-opus-5-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "constraints": { + "thinkingOnly": true + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Opus 5.5", + "description": "Claude 订阅额度计费，长程智能体编程与知识工作模型，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic-claude.json#claude-opus-5-5" + }, + "requiredClientVersion": "10.0.83" + }, + { + "apiModelId": "claude-fable-5-1", + "modelRef": "anthropic:claude-fable-5-1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic-claude.json#claude-fable-5-1", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "constraints": { + "thinkingOnly": true + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Fable 5.1", + "description": "Claude 订阅额度计费，面向高难推理和长程智能体任务，1M 上下文，始终开启自适应思考", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic-claude.json#claude-fable-5-1" + }, + "requiredClientVersion": "10.0.83" + }, + { + "apiModelId": "claude-opus-5", + "modelRef": "anthropic:claude-opus-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic-claude.json#claude-opus-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Opus 5", + "description": "Claude 订阅额度计费，Opus 5 系列模型，复杂智能体编码与长程任务首选，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic-claude.json#claude-opus-5" + }, + "requiredClientVersion": "10.0.83" + }, + { + "apiModelId": "claude-sonnet-5-5", + "modelRef": "anthropic:claude-sonnet-5-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic-claude.json#claude-sonnet-5-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "constraints": { + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Sonnet 5.5", + "description": "Claude Sonnet 5.5，速度与智能平衡，1M 上下文，默认 high effort，自适应思考", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic-claude.json#claude-sonnet-5-5" + }, + "requiredClientVersion": "10.0.83" + }, + { + "apiModelId": "claude-sonnet-5", + "modelRef": "anthropic:claude-sonnet-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic-claude.json#claude-sonnet-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless" + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Sonnet 5", + "description": "Claude 订阅额度计费，速度与智能平衡模型，支持长上下文智能体任务", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic-claude.json#claude-sonnet-5" + }, + "requiredClientVersion": "10.0.83" + } + ], + "metadata": { + "sourceFile": "compute/providers/anthropic-claude.json" + } + }, + { + "id": "provider-anthropic-001", + "category": "providers", + "connection": { + "id": "provider-anthropic-001", + "provider": "anthropic", + "label": "Anthropic Claude", + "baseUrl": "https://api.anthropic.com", + "apiFormat": "anthropic-messages", + "apiKeyRef": "anthropic", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "computer_use" + ], + "tombstones": [ + "claude-opus-4-8", + "claude-opus-4-7", + "claude-sonnet-4-6", + "claude-fable-5", + "claude-haiku-4-5", + "claude-haiku-4-5-20251001" + ] + }, + "models": [ + { + "apiModelId": "claude-opus-5-5", + "modelRef": "anthropic:claude-opus-5-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-opus-5-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 2, + "toolType": "web_search_20260318" + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "constraints": { + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true, + "thinkingOnly": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "write5m": 5, + "write1h": 8, + "read": 0.2 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Full 1M context is billed at standard pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Opus 5.5", + "description": "Anthropic 面向长程智能体编程与知识工作的 Opus 模型，1M 上下文，始终开启自适应思考" + }, + "recordedExtra": { + "promptCacheMinTokens": 512 + }, + "sourceRecord": "compute/providers/anthropic.json#claude-opus-5-5" + } + }, + { + "apiModelId": "claude-fable-5-1", + "modelRef": "anthropic:claude-fable-5-1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-fable-5-1", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 1, + "toolType": "web_search_20260318" + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "constraints": { + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true, + "thinkingOnly": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 50, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "write5m": 12.5, + "write1h": 20, + "read": 0.25 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Full 1M context is billed at standard pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Fable 5.1", + "description": "面向高难推理和长程智能体任务，1M 上下文，始终开启自适应思考" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic.json#claude-fable-5-1" + } + }, + { + "apiModelId": "claude-opus-5", + "modelRef": "anthropic:claude-opus-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-opus-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 2, + "toolType": "web_search_20260318" + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "constraints": { + "samplingParametersDeprecated": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 25, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "write5m": 6.25, + "write1h": 10, + "read": 0.5 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Full 1M context is billed at standard pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Opus 5", + "description": "Anthropic Opus 5 系列模型，复杂智能体编码与长程任务能力显著强于 Opus 4.8，价格不变，1M 上下文" + }, + "recordedExtra": { + "promptCacheMinTokens": 512 + }, + "sourceRecord": "compute/providers/anthropic.json#claude-opus-5" + } + }, + { + "apiModelId": "claude-sonnet-5-5", + "modelRef": "anthropic:claude-sonnet-5-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "computer_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-sonnet-5-5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-sonnet-5-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 3, + "toolType": "web_search_20260318" + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "constraints": { + "samplingParametersDeprecated": true, + "forcedToolChoiceUnsupported": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "write5m": 2.5, + "write1h": 4, + "read": 0.2 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Sonnet 5.5", + "description": "Claude Sonnet 5.5，速度与智能平衡，1M 上下文，默认 high effort，自适应思考" + }, + "recordedExtra": { + "promptCacheMinTokens": 512 + }, + "sourceRecord": "compute/providers/anthropic.json#claude-sonnet-5-5" + } + }, + { + "apiModelId": "claude-sonnet-5", + "modelRef": "anthropic:claude-sonnet-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "computer_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-sonnet-5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "computer_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/anthropic.json#claude-sonnet-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinkingRoundTrip": { + "enabled": true, + "protocol": "anthropic-messages", + "allowEmptySignature": false, + "scope": "active-tool-turn" + }, + "thinkingToolTurnValidation": "accepts-thinkless", + "serverSideWebSearch": { + "enabled": true, + "dialect": "anthropic-messages", + "fallbackPriority": 3, + "toolType": "web_search_20260318" + }, + "adaptiveThinking": true, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "high" + }, + "constraints": { + "samplingParametersDeprecated": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "write5m": 2.5, + "write1h": 4, + "read": 0.2 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Claude Sonnet 5", + "description": "Anthropic Sonnet 5，速度与智能平衡模型，支持电脑使用和长上下文智能体任务" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/anthropic.json#claude-sonnet-5" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/anthropic.json" + } + }, + { + "id": "provider-baichuan-001", + "category": "providers", + "connection": { + "id": "provider-baichuan-001", + "provider": "baichuan", + "label": "百川 Baichuan", + "baseUrl": "https://api.baichuan-ai.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "baichuan", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat" + ] + }, + "models": [ + { + "apiModelId": "Baichuan-M3-Plus", + "modelRef": "baichuan:Baichuan-M3-Plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.85, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baichuan.json#Baichuan-M3-Plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 9, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "百川 M3-Plus", + "description": "百川智能M3-Plus医疗增强模型，235B参数，幻觉率仅2.6%，API价格较前代下降70%" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baichuan.json#Baichuan-M3-Plus" + } + }, + { + "apiModelId": "Baichuan-M3", + "modelRef": "baichuan:Baichuan-M3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.85, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baichuan.json#Baichuan-M3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "百川 M3", + "description": "百川智能M3新一代开源医疗增强模型，235B参数" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baichuan.json#Baichuan-M3" + } + }, + { + "apiModelId": "Baichuan-M2-Plus", + "modelRef": "baichuan:Baichuan-M2-Plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.85, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baichuan.json#Baichuan-M2-Plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "百川 M2-Plus", + "description": "百川智能M2-Plus旗舰模型，最强推理和创作能力" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baichuan.json#Baichuan-M2-Plus" + } + }, + { + "apiModelId": "Baichuan-M2", + "modelRef": "baichuan:Baichuan-M2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.85, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baichuan.json#Baichuan-M2", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "百川 M2", + "description": "百川智能M2模型，高性能通用模型" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baichuan.json#Baichuan-M2" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/baichuan.json" + } + }, + { + "id": "provider-baidu-001", + "category": "providers", + "connection": { + "id": "provider-baidu-001", + "provider": "baidu", + "label": "百度 文心", + "baseUrl": "https://qianfan.baidubce.com/v2", + "apiFormat": "openai-completions", + "apiKeyRef": "baidu", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ] + }, + "models": [ + { + "apiModelId": "ernie-5.0-thinking-latest", + "modelRef": "baidu:ernie-5.0-thinking-latest", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 24, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32768, + "inputPrice": 6, + "outputPrice": 24 + }, + { + "maxInputTokens": 131072, + "inputPrice": 10, + "outputPrice": 40 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "文心 ERNIE 5.0", + "description": "百度最新旗舰模型，支持深度推理、多模态和工具调用，119K输入/64K输出" + }, + "recordedExtra": { + "thinkingMaxTokens": 60000 + }, + "sourceRecord": "compute/providers/baidu.json#ernie-5.0-thinking-latest" + } + }, + { + "apiModelId": "ernie-5.0", + "modelRef": "baidu:ernie-5.0", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baidu.json#ernie-5.0", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 24, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32768, + "inputPrice": 6, + "outputPrice": 24 + }, + { + "maxInputTokens": 131072, + "inputPrice": 10, + "outputPrice": 40 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "文心 ERNIE 5.0 非思考", + "description": "百度 ERNIE 5.0 旗舰模型，128K 上下文，119K 最大输入，最大输出 65,536 token" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baidu.json#ernie-5.0" + } + }, + { + "apiModelId": "ernie-4.5-turbo-128k", + "modelRef": "baidu:ernie-4.5-turbo-128k", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baidu.json#ernie-4.5-turbo-128k", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "文心 ERNIE 4.5 Turbo", + "description": "百度高性价比长上下文模型，128K 窗口，最大输出 12,288 token" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baidu.json#ernie-4.5-turbo-128k" + } + }, + { + "apiModelId": "ernie-4.5-turbo-20260402", + "modelRef": "baidu:ernie-4.5-turbo-20260402", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/baidu.json#ernie-4.5-turbo-20260402", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "文心 ERNIE 4.5 Turbo 20260402", + "description": "百度 ERNIE 4.5 Turbo 20260402 快照，128K 上下文，最大输出 12,288 token" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/baidu.json#ernie-4.5-turbo-20260402" + } + }, + { + "apiModelId": "ernie-x1.1", + "modelRef": "baidu:ernie-x1.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "文心 ERNIE X1.1", + "description": "百度 ERNIE X1.1 深度思考模型，64K 上下文，最大输出 65,536 token" + }, + "recordedExtra": { + "thinkingMaxTokens": 65536 + }, + "sourceRecord": "compute/providers/baidu.json#ernie-x1.1" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/baidu.json" + } + }, + { + "id": "provider-cohere-001", + "category": "providers", + "connection": { + "id": "provider-cohere-001", + "provider": "cohere", + "label": "Cohere", + "baseUrl": "https://api.cohere.ai/compatibility/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "cohere", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "embedding", + "rerank" + ], + "tombstones": [ + "rerank-v3.5" + ] + }, + "models": [ + { + "apiModelId": "embed-v5.0-fast", + "modelRef": "cohere:embed-v5.0-fast", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "text_embedding", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/cohere.json#embed-v5.0-fast", + "verification": "inherited-not-reverified" + } + }, + "request": { + "constraints": { + "defaultDimension": 2048 + } + }, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "Token 单价未由当前官方价格页公开确认；Model Vault 按实例计价。 Compatibility API 不接受 dimensions 参数；返回默认维度，多维度选择需原生 Embed API。", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Embed 5 Fast", + "description": "Embed 5 Fast；当前 OpenAI 兼容接入仅声明文本向量能力" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/cohere.json#embed-v5.0-fast" + } + }, + { + "apiModelId": "embed-v5.0-pro", + "modelRef": "cohere:embed-v5.0-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "text_embedding", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/cohere.json#embed-v5.0-pro", + "verification": "inherited-not-reverified" + } + }, + "request": { + "constraints": { + "defaultDimension": 2048 + } + }, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "Token 单价未由当前官方价格页公开确认；Model Vault 按实例计价。 Compatibility API 不接受 dimensions 参数；返回默认维度，多维度选择需原生 Embed API。", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Embed 5 Pro", + "description": "Embed 5 Pro；当前 OpenAI 兼容接入仅声明文本向量能力" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/cohere.json#embed-v5.0-pro" + } + }, + { + "apiModelId": "command-a-plus-05-2026", + "modelRef": "cohere:command-a-plus-05-2026", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "none", + "high" + ] + } + }, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "官方文档声明限流额度内免费；生产 Model Vault 按实例单独计价，不预填 token 单价。", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Command A+", + "description": "Command A+" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/cohere.json#command-a-plus-05-2026" + } + }, + { + "apiModelId": "command-a-03-2025", + "modelRef": "cohere:command-a-03-2025", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.75, + "reason": "preserved-access-contract", + "evidence": "compute/providers/cohere.json#command-a-03-2025", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Command A", + "description": "Cohere Command A 文本模型，256K 上下文，支持 RAG 和工具调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/cohere.json#command-a-03-2025" + } + }, + { + "apiModelId": "command-r7b-12-2024", + "modelRef": "cohere:command-r7b-12-2024", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.75, + "reason": "preserved-access-contract", + "evidence": "compute/providers/cohere.json#command-r7b-12-2024", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.0375, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.15, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Command R7B", + "description": "Cohere 小型高速对话模型，适合高吞吐、低延迟场景" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/cohere.json#command-r7b-12-2024" + } + }, + { + "apiModelId": "embed-v4.0", + "modelRef": "cohere:embed-v4.0", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "maxOutputTokens": { + "value": 0, + "reason": "preserved-access-contract", + "evidence": "compute/providers/cohere.json#embed-v4.0", + "verification": "inherited-not-reverified" + } + }, + "request": { + "constraints": { + "defaultDimension": 1536 + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.12, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Embedding models are priced by embedded tokens; price is per 1M tokens. Compatibility API 不接受 dimensions 参数；返回默认维度，多维度选择需原生 Embed API。", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Embed V4", + "description": "Cohere 最新 Embedding 模型，128K 上下文，多语言支持" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/cohere.json#embed-v4.0" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/cohere.json" + } + }, + { + "id": "provider-dashscope-001", + "category": "providers", + "connection": { + "id": "provider-dashscope-001", + "provider": "dashscope", + "label": "阿里 通义千问", + "baseUrl": "https://dashscope.aliyuncs.com/compatible-mode/v1", + "mediaBaseUrl": "https://dashscope.aliyuncs.com/api/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "dashscope", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "embedding", + "rerank", + "tts", + "asr", + "vision", + "image_gen", + "video_gen", + "voice_clone", + "omni", + "translation" + ], + "tombstones": [ + "qwen3.8-max-preview", + "qwen3.6-plus", + "qwen3.5-plus", + "qwen-max", + "qwen-plus", + "qwen-turbo", + "qwen3-max", + "qwen3-max-trans" + ] + }, + "models": [ + { + "apiModelId": "qwen-image-3.0", + "modelRef": "qwen:qwen-image-3.0", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "按图片计费，北京：输入 0.02 元/张；Pro 输出 1K 0.25 元、2K 0.5 元；普通版输出 0.18 元/张。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "qwen-image-3.0", + "description": "qwen-image-3.0" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen-image-3.0", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "qwen-image-3.0-pro", + "modelRef": "qwen:qwen-image-3.0-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "按图片计费，北京：输入 0.02 元/张；Pro 输出 1K 0.25 元、2K 0.5 元；普通版输出 0.18 元/张。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "qwen-image-3.0-pro", + "description": "qwen-image-3.0-pro" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen-image-3.0-pro", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "qwen3.7-text-rerank", + "modelRef": "qwen:qwen3.7-text-rerank", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.7 Text Rerank", + "description": "Qwen3.7 Text Rerank" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3.7-text-rerank" + } + }, + { + "apiModelId": "qwen3.7-text-embedding-flash", + "modelRef": "qwen:qwen3.7-text-embedding-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.125, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "qwen3.7-text-embedding-flash", + "description": "qwen3.7-text-embedding-flash" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3.7-text-embedding-flash" + } + }, + { + "apiModelId": "qwen3.7-text-embedding", + "modelRef": "qwen:qwen3.7-text-embedding", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "qwen3.7-text-embedding", + "description": "qwen3.7-text-embedding" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3.7-text-embedding" + } + }, + { + "apiModelId": "qwen3.8-omni-flash", + "modelRef": "qwen:qwen3.8-omni-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.7, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.1, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.8 Omni Flash", + "description": "Qwen3.8 Omni Flash" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3.8-omni-flash" + } + }, + { + "apiModelId": "qwen3.8-max", + "modelRef": "qwen:qwen3.8-max", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.8-max", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision", + "video_understanding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.8-max", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 12, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 36, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 1.5, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "CNY per 1M tokens, Beijing region standard price.", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3.8-Max", + "description": "通义千问3.8 Max 稳定版，支持视觉、视频理解和工具调用，100万上下文" + }, + "recordedExtra": { + "preserveThinkingDefault": true, + "thinkingMaxTokens": 262144 + }, + "sourceRecord": "compute/providers/dashscope.json#qwen3.8-max" + } + }, + { + "apiModelId": "qwen3.8-flash", + "modelRef": "qwen:qwen3.8-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.8-flash", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision", + "video_understanding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.8-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.7, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.1, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "CNY per 1M tokens, Beijing region standard price.", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3.8-Flash", + "description": "通义千问3.8 Flash 稳定版，高速多模态推理，100万上下文" + }, + "recordedExtra": { + "preserveThinkingDefault": true, + "thinkingMaxTokens": 262144 + }, + "sourceRecord": "compute/providers/dashscope.json#qwen3.8-flash" + } + }, + { + "apiModelId": "qwen3.7-max", + "modelRef": "qwen:qwen3.7-max", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-max", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 0.6, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-max", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-max", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 12, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 36, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 2.4, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3.7-Max", + "description": "通义千问3.7 Max，面向复杂推理与长程 Agent 任务的旗舰模型，100万上下文，支持内置工具和 Function Calling" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3.7-max" + } + }, + { + "apiModelId": "qwen3.7-plus", + "modelRef": "qwen:qwen3.7-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-plus", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "long_context", + "tool_use", + "agent", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-plus", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 0.6, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-plus", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.7-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 256000, + "inputPrice": 2, + "outputPrice": 8, + "thinkingOutputPrice": 8 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 6, + "outputPrice": 24, + "thinkingOutputPrice": 24 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3.7-Plus", + "description": "通义千问3.7 Plus，多模态智能体模型，100万上下文，支持内置工具、Function Calling 和结构化输出" + }, + "recordedExtra": { + "thinkingMaxTokens": 262144 + }, + "sourceRecord": "compute/providers/dashscope.json#qwen3.7-plus" + } + }, + { + "apiModelId": "qwen3.6-flash", + "modelRef": "qwen:qwen3.6-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3.6-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 7.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 256000, + "inputPrice": 1.2, + "outputPrice": 7.2 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 4.8, + "outputPrice": 28.8 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3.6-Flash", + "description": "通义千问3.6 Flash，高性价比多模态模型，100万上下文，支持内置工具和 Function Calling" + }, + "recordedExtra": { + "thinkingMaxTokens": 131072 + }, + "sourceRecord": "compute/providers/dashscope.json#qwen3.6-flash" + } + }, + { + "apiModelId": "qwen-long", + "modelRef": "qwen:qwen-long", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.8, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen-long", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 qwen-long", + "description": "通义千问Long，支持1000万上下文，适合超长文档处理，降价97%" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen-long" + } + }, + { + "apiModelId": "qwen3-vl-plus", + "modelRef": "qwen:qwen3-vl-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.8, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3-vl-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3-VL-Plus", + "description": "通义千问3代视觉模型Plus版，支持图像理解" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3-vl-plus" + } + }, + { + "apiModelId": "qwen3-vl-flash", + "modelRef": "qwen:qwen3-vl-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.8, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen3-vl-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 Qwen3-VL-Flash", + "description": "通义千问3代视觉模型Flash版，高性价比" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3-vl-flash" + } + }, + { + "apiModelId": "text-embedding-v3", + "modelRef": "qwen:text-embedding-v3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.7, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "text-embedding-v3", + "description": "阿里云文本向量化" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#text-embedding-v3" + } + }, + { + "apiModelId": "text-embedding-v4", + "modelRef": "qwen:text-embedding-v4", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "阿里云 text-embedding-v4", + "description": "阿里云百炼文本向量模型v4，支持多维度与多语种" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#text-embedding-v4" + } + }, + { + "apiModelId": "qwen3-rerank", + "modelRef": "qwen:qwen3-rerank", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3-Rerank", + "description": "检索重排序，支持 100+ 语种" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen3-rerank" + } + }, + { + "apiModelId": "cosyvoice-v2", + "modelRef": "qwen:cosyvoice-v2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "CosyVoice V2", + "description": "通义语音合成，支持中英文" + }, + "recordedExtra": { + "voices": [ + "longxiaochun", + "longxiaoxia", + "longlaotie", + "longshu" + ], + "supportedAudioFormats": [ + "mp3", + "wav", + "pcm" + ] + }, + "sourceRecord": "compute/providers/dashscope.json#cosyvoice-v2" + } + }, + { + "apiModelId": "paraformer-v2", + "modelRef": "qwen:paraformer-v2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Paraformer V2", + "description": "通义语音识别" + }, + "recordedExtra": { + "maxAudioLength": 3600, + "supportedInputFormats": [ + "wav", + "mp3", + "pcm", + "ogg" + ], + "streamingSupported": true + }, + "sourceRecord": "compute/providers/dashscope.json#paraformer-v2", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/asr-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:asr:dashscope" + } + }, + { + "apiModelId": "wan2.7-image-pro", + "modelRef": "qwen:wan2.7-image-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.7 Pro", + "description": "万相最新旗舰文生图模型，支持 4K 分辨率、thinking_mode 智能思考、色彩调色板控制" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "768*768", + "1024*1024", + "1280*1280", + "1024*1792", + "1792*1024", + "2048*2048", + "4096*4096" + ], + "sizeShortcuts": { + "1K": "1024*1024", + "2K": "2048*2048", + "4K": "4096*4096" + }, + "maxN": 4, + "supportsThinkingMode": true, + "supportsColorPalette": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan2.7-image-pro", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "wan2.7-image", + "modelRef": "qwen:wan2.7-image", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.7", + "description": "万相最新标准文生图模型，高画质，支持 thinking_mode" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "768*768", + "1024*1024", + "1280*1280", + "1024*1792", + "1792*1024" + ], + "maxN": 4, + "supportsThinkingMode": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan2.7-image", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "wan2.6-t2i", + "modelRef": "qwen:wan2.6-t2i", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.6 文生图", + "description": "万相 2.6 文生图专用模型，稳定可靠，支持 prompt_extend 智能改写" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "1024*1024", + "1280*1280", + "720*1280", + "1280*720", + "768*1152", + "1152*768" + ], + "maxN": 4, + "supportsNegativePrompt": true, + "supportsPromptExtend": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan2.6-t2i", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "wan2.2-t2i-plus", + "modelRef": "qwen:wan2.2-t2i-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.2 Plus", + "description": "万相 2.2 增强版，画质与速度平衡" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "1024*1024", + "720*1280", + "1280*720" + ], + "maxN": 4, + "supportsNegativePrompt": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan2.2-t2i-plus", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "wan2.2-t2i-flash", + "modelRef": "qwen:wan2.2-t2i-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.2 Flash", + "description": "万相 2.2 快速版，生成速度极快，适合批量和实时场景" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "1024*1024", + "720*1280", + "1280*720" + ], + "maxN": 4, + "supportsNegativePrompt": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan2.2-t2i-flash", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "qwen-image-2.0-pro", + "modelRef": "qwen:qwen-image-2.0-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "千问图像 2.0 Pro", + "description": "千问图像生成 2.0 Pro 旗舰版，高画质多风格" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "512*512", + "1024*1024", + "1280*1280", + "1664*928", + "928*1664" + ], + "maxN": 4 + }, + "sourceRecord": "compute/providers/dashscope.json#qwen-image-2.0-pro", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "qwen-image-2.0", + "modelRef": "qwen:qwen-image-2.0", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "千问图像 2.0", + "description": "千问图像生成 2.0 标准版" + }, + "recordedExtra": { + "endpoint": "/services/aigc/image-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "supportedImageSizes": [ + "512*512", + "1024*1024", + "1280*1280", + "2048*2048" + ], + "maxN": 4 + }, + "sourceRecord": "compute/providers/dashscope.json#qwen-image-2.0", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "wan3.0-video-prime", + "modelRef": "wan:wan3.0-video-prime", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "chinese_optimized", + "high_quality", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#wan3.0-video-prime", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 3.0 视频生成高速版", + "description": "万相 3.0 全能参考视频生成高速版，能力对齐标准版并显著提升端到端生成速度" + }, + "recordedExtra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 30, + "supportsIntelligentDuration": true, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ], + "supportedRatios": [ + "adaptive", + "16:9", + "4:3", + "1:1", + "3:4", + "9:16" + ], + "mediaInputTypes": [ + "first_frame", + "last_frame", + "reference_image", + "reference_video", + "reference_audio", + "file", + "link" + ], + "maxReferenceImages": 10, + "maxReferenceVideos": 5, + "maxReferenceAudios": 5, + "supportsOutputAudio": true, + "supportsSeed": true, + "supportsPromptExtend": true, + "supportsWatermark": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan3.0-video-prime" + } + }, + { + "apiModelId": "wan3.0-video", + "modelRef": "wan:wan3.0-video", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "text_to_video", + "image_to_video", + "reference_to_video", + "multi_image", + "reference_video", + "reference_audio", + "first_last_frame", + "audio_generation", + "chinese_optimized", + "high_quality" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#wan3.0-video", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 3.0 视频生成", + "description": "万相 3.0 全能参考视频生成标准版，统一支持文生、首帧/首尾帧图生及多类型参考生视频" + }, + "recordedExtra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 30, + "supportsIntelligentDuration": true, + "fps": 30, + "supportedResolutions": [ + "480P", + "720P", + "1080P" + ], + "supportedRatios": [ + "adaptive", + "16:9", + "4:3", + "1:1", + "3:4", + "9:16" + ], + "mediaInputTypes": [ + "first_frame", + "last_frame", + "reference_image", + "reference_video", + "reference_audio", + "file", + "link" + ], + "maxReferenceImages": 10, + "maxReferenceVideos": 5, + "maxReferenceAudios": 5, + "supportsOutputAudio": true, + "supportsSeed": true, + "supportsPromptExtend": true, + "supportsWatermark": true + }, + "sourceRecord": "compute/providers/dashscope.json#wan3.0-video" + } + }, + { + "apiModelId": "wan2.6-t2v", + "modelRef": "qwen:wan2.6-t2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.6 文生视频", + "description": "万相 2.6 文生视频模型，支持文生视频和图生视频" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "720p", + "1080p" + ], + "endpoint": "/services/aigc/video-generation/generation", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks" + }, + "sourceRecord": "compute/providers/dashscope.json#wan2.6-t2v" + } + }, + { + "apiModelId": "cosyvoice-clone", + "modelRef": "qwen:cosyvoice-clone", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "CosyVoice 声音复刻", + "description": "通义声音复刻，少量样本即可克隆" + }, + "recordedExtra": { + "maxCloneSamples": 5, + "minCloneDuration": 3 + }, + "sourceRecord": "compute/providers/dashscope.json#cosyvoice-clone" + } + }, + { + "apiModelId": "qwen-omni-turbo", + "modelRef": "qwen:qwen-omni-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.01, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen-omni-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen-Omni-Turbo", + "description": "通义全能模型，支持音视频文本多模态" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/dashscope.json#qwen-omni-turbo" + } + }, + { + "apiModelId": "qwen-mt-plus", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "translation" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen-mt-plus", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 16384, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen-mt-plus", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen-mt-plus", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "translation", + "multilingual", + "chinese_optimized", + "terminology" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#qwen-mt-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 5.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen-MT-Plus", + "description": "通义千问 Qwen-MT 旗舰翻译模型，支持 92 种语言互译，适合专业领域、正式文书和技术报告" + }, + "recordedExtra": { + "supportedLanguageCount": 92, + "supportsIncrementalStreaming": false + }, + "sourceRecord": "compute/providers/dashscope.json#qwen-mt-plus", + "automaticRouting": false + } + }, + { + "apiModelId": "happyhorse-1.0-t2v", + "modelRef": "happyhorse:happyhorse-1.0-t2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "text_to_video", + "chinese_optimized", + "high_quality" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#happyhorse-1.0-t2v", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "快乐马 1.0 文生视频", + "description": "阿里 HappyHorse 文生视频，原生音视频联合生成，最长 15s，支持多语种口型" + }, + "recordedExtra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 15, + "supportedResolutions": [ + "720P", + "1080P" + ] + }, + "sourceRecord": "compute/providers/dashscope.json#happyhorse-1.0-t2v" + } + }, + { + "apiModelId": "happyhorse-1.0-i2v", + "modelRef": "happyhorse:happyhorse-1.0-i2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "image_to_video", + "chinese_optimized" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#happyhorse-1.0-i2v", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "快乐马 1.0 图生视频", + "description": "阿里 HappyHorse 图生视频（首帧），宽高比自动跟随输入首帧" + }, + "recordedExtra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxVideoDuration": 15, + "supportedResolutions": [ + "720P", + "1080P" + ] + }, + "sourceRecord": "compute/providers/dashscope.json#happyhorse-1.0-i2v" + } + }, + { + "apiModelId": "happyhorse-1.0-r2v", + "modelRef": "happyhorse:happyhorse-1.0-r2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "reference_to_video", + "multi_image", + "chinese_optimized" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/dashscope.json#happyhorse-1.0-r2v", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "快乐马 1.0 参考生视频", + "description": "阿里 HappyHorse 参考生视频，支持最多 9 张参考图（多主体/元素一致性）" + }, + "recordedExtra": { + "endpoint": "/services/aigc/video-generation/video-synthesis", + "asyncHeader": "X-DashScope-Async", + "taskQueryEndpoint": "/tasks", + "maxReferenceImages": 9, + "maxVideoDuration": 15, + "supportedResolutions": [ + "720P", + "1080P" + ] + }, + "sourceRecord": "compute/providers/dashscope.json#happyhorse-1.0-r2v" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/dashscope.json" + } + }, + { + "id": "provider-deepseek-001", + "category": "providers", + "connection": { + "id": "provider-deepseek-001", + "provider": "deepseek", + "label": "DeepSeek", + "baseUrl": "https://api.deepseek.com", + "apiFormat": "openai-completions", + "apiKeyRef": "deepseek", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning", + "vision" + ], + "tombstones": [ + "deepseek-chat", + "deepseek-reasoner" + ] + }, + "models": [ + { + "apiModelId": "deepseek-flash", + "modelRef": "deepseek:deepseek-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "long_context", + "multilingual", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-flash", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.02, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "idle", + "inputPrice": 1, + "outputPrice": 4, + "cacheHitPrice": 0.02 + }, + { + "condition": "peak", + "inputPrice": 2, + "outputPrice": 8, + "cacheHitPrice": 0.04 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "空闲时段为北京时间工作日 09:00-12:00、14:00-18:00 之外的时段；高峰时段价格为输入 2 元、输出 8 元、缓存命中输入 0.04 元/百万 tokens。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4.1 Flash", + "description": "DeepSeek V4.1 Flash，支持文本与图像输入，1M 上下文，最大输出 384K，并发上限 2500" + }, + "recordedExtra": { + "concurrencyLimit": 2500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": "non-thinking-only" + } + }, + "sourceRecord": "compute/providers/deepseek.json#deepseek-flash" + } + }, + { + "apiModelId": "deepseek-v4-flash-vision-exp", + "modelRef": "deepseek:deepseek-v4-flash-vision-exp", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "vision", + "long_context", + "multilingual", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-v4-flash-vision-exp", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-v4-flash-vision-exp", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.008, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.016, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.02016, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Flash Vision Exp", + "description": "DeepSeek V4 Flash 实验性视觉版本，支持文本与图像输入，文本能力与 V4 Flash 一致，1M 上下文，并发上限 2500" + }, + "recordedExtra": { + "concurrencyLimit": 2500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": false + } + }, + "sourceRecord": "compute/providers/deepseek.json#deepseek-v4-flash-vision-exp" + } + }, + { + "apiModelId": "deepseek-v4-flash", + "modelRef": "deepseek:deepseek-v4-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "long_context", + "multilingual", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-v4-flash", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-v4-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.008, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.016, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.02016, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Flash", + "description": "DeepSeek V4 Flash，284B 总参数 / 13B 激活 MoE，1M 上下文，默认思考模式，支持非思考模式，并发上限 2500" + }, + "recordedExtra": { + "concurrencyLimit": 2500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": "non-thinking-only" + } + }, + "sourceRecord": "compute/providers/deepseek.json#deepseek-v4-flash" + } + }, + { + "apiModelId": "deepseek-v4-pro", + "modelRef": "deepseek:deepseek-v4-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "code", + "reasoning", + "deep_thinking", + "long_context", + "multilingual", + "tool_use", + "math" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-v4-pro", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/deepseek.json#deepseek-v4-pro", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3.132, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 6.264, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.0261, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Pro", + "description": "DeepSeek V4 Pro，1.6T 总参数 / 49B 激活 MoE，1M 上下文，默认思考模式，支持非思考模式，并发上限 500" + }, + "recordedExtra": { + "concurrencyLimit": 500, + "supportsThinking": true, + "thinkingDefault": true, + "features": { + "jsonOutput": true, + "toolCalls": true, + "chatPrefixCompletion": "beta", + "fimCompletion": "non-thinking-only" + } + }, + "sourceRecord": "compute/providers/deepseek.json#deepseek-v4-pro" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/deepseek.json" + } + }, + { + "id": "provider-github-copilot-plan-001", + "category": "providers", + "connection": { + "id": "provider-github-copilot-plan-001", + "provider": "github-copilot", + "brandGroup": "github", + "label": "GitHub Copilot 订阅", + "baseUrl": "https://api.githubcopilot.com", + "apiFormat": "openai-completions", + "apiKeyRef": "github-copilot-session-token", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "accessMode": "coding-plan", + "credentialSource": "github-copilot", + "requiredClientVersion": "10.0.132", + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://github.com/settings/copilot/features" + } + }, + "services": [ + "chat", + "reasoning" + ] + }, + "models": [], + "metadata": { + "sourceFile": "compute/providers/github-copilot.json" + } + }, + { + "id": "provider-google-001", + "category": "providers", + "connection": { + "id": "provider-google-001", + "provider": "google", + "label": "Google Gemini", + "baseUrl": "https://generativelanguage.googleapis.com/v1beta", + "apiFormat": "google-generative-ai", + "apiKeyRef": "google", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "vision", + "embedding" + ], + "tombstones": [ + "gemini-3.1-flash-lite-preview" + ] + }, + "models": [ + { + "apiModelId": "gemini-embedding-2", + "modelRef": "google:gemini-embedding-2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "constraints": { + "dimensions": [ + 768, + 1536, + 3072 + ], + "defaultDimension": 3072 + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini Embedding 2", + "description": "Gemini Embedding 2，原生多模态嵌入，8192 token 输入，共享文本/图像/音视频向量空间" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-embedding-2" + } + }, + { + "apiModelId": "gemini-3.8-flash", + "modelRef": "google:gemini-3.8-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.8-flash", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.8-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high" + ], + "defaultEffort": "medium" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.075 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Standard paid tier introductory price through 2026-12-31; standard rates rise on 2027-01-01.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3.8 Flash", + "description": "Google 最新稳定版 Flash，面向长程编程和智能体任务，1M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3.8-flash" + } + }, + { + "apiModelId": "gemini-3.7-flash", + "modelRef": "google:gemini-3.7-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.7-flash", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.7-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high" + ], + "defaultEffort": "medium" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.075 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Standard paid tier introductory price through 2026-12-31; standard rates rise on 2027-01-01.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3.7 Flash", + "description": "Google 稳定版 Flash，面向日常编程和多步工具调用，1M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3.7-flash" + } + }, + { + "apiModelId": "gemini-3.5-flash-lite", + "modelRef": "google:gemini-3.5-flash-lite", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.5-flash-lite", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.5-flash-lite", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.5-flash-lite", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.03 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3.5 Flash-Lite", + "description": "Google 低延迟轻量模型，支持多模态输入，1M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3.5-flash-lite" + } + }, + { + "apiModelId": "gemini-3.5-flash", + "modelRef": "google:gemini-3.5-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.5-flash", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.5-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 9, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.75 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Standard paid tier text/image/video price; audio input uses a separate Google pricing row.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3.5 Flash", + "description": "Google Gemini 3.5 Flash 稳定版，高速多模态模型，1M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3.5-flash" + } + }, + { + "apiModelId": "gemini-3.1-flash-lite", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-flash-lite", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1048576, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-flash-lite", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 65536, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-flash-lite", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-flash-lite", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-flash-lite", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-flash-lite", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioInputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.125, + "audioInputCacheRead": 0.25 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3.1 Flash-Lite", + "description": "Google Gemini 3.1 Flash-Lite 稳定版，低成本高吞吐多模态模型，替代已关闭的 Preview 版本" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3.1-flash-lite", + "automaticRouting": false + } + }, + { + "apiModelId": "gemini-3.1-pro-preview", + "modelRef": "google:gemini-3.1-pro-preview", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3.1-pro-preview", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 12, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Standard paid tier text/image/video price for prompts <= 200K tokens; prompts > 200K are $4 input and $18 output per 1M tokens.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.2, + "inputCacheReadOver200k": 0.4, + "storagePerMillionTokensPerHour": 4.5 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3.1 Pro Preview", + "description": "Google Gemini 3.1 Pro Preview，面向复杂推理、编码和智能体工作流" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3.1-pro-preview" + } + }, + { + "apiModelId": "gemini-3-flash-preview", + "modelRef": "google:gemini-3-flash-preview", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-3-flash-preview", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Standard paid tier text/image/video price; audio input is $1.00 per 1M tokens.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachePricing", + "amount": { + "inputCacheRead": 0.05, + "audioInputCacheRead": 0.1, + "storagePerMillionTokensPerHour": 1 + }, + "currency": "USD", + "unit": "compound-source-units", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 3 Flash Preview", + "description": "Google Gemini 3 Flash Preview，高速多模态推理模型" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-3-flash-preview" + } + }, + { + "apiModelId": "gemini-2.5-pro", + "modelRef": "google:gemini-2.5-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-2.5-pro", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 2.5 Pro", + "description": "Google Gemini 2.5 Pro，1,048,576 token 上下文，多模态" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-2.5-pro" + } + }, + { + "apiModelId": "gemini-2.5-flash", + "modelRef": "google:gemini-2.5-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/google.json#gemini-2.5-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Gemini 2.5 Flash", + "description": "Google 高性价比模型，1,048,576 token 上下文，速度快" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#gemini-2.5-flash" + } + }, + { + "apiModelId": "text-embedding-005", + "modelRef": "google:text-embedding-005", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Text Embedding 005", + "description": "Google 文本嵌入模型，768维度" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/google.json#text-embedding-005" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/google.json" + } + }, + { + "id": "provider-kling-001", + "category": "providers", + "connection": { + "id": "provider-kling-001", + "provider": "kling", + "label": "快手 可灵", + "baseUrl": "https://api.klingai.com/v1", + "apiFormat": "kling-task-api", + "apiKeyRef": "kling", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "video_gen" + ] + }, + "models": [ + { + "apiModelId": "kling-v3", + "modelRef": "kling:kling-v3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "可灵 3.0", + "description": "可灵 3.0，官方 legacy /v1/videos/text2video 协议已支持该 model_name" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/kling.json#kling-v3", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-video.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-video_gen-provider-adapter", + "apiFormat": "kling-task-api", + "adapterId": "required-adapter:video_gen:kling" + } + }, + { + "apiModelId": "kling-v2-5-turbo", + "modelRef": "kling:kling-v2-5-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerGeneration", + "amount": 1.5, + "currency": "CNY", + "unit": "generation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "可灵 V2.5 Turbo", + "description": "可灵最新视频生成模型，高性价比" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "720p", + "1080p" + ] + }, + "sourceRecord": "compute/providers/kling.json#kling-v2-5-turbo", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-video.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-video_gen-provider-adapter", + "apiFormat": "kling-task-api", + "adapterId": "required-adapter:video_gen:kling" + } + }, + { + "apiModelId": "kling-v2-5-turbo-pro", + "modelRef": "kling:kling-v2-5-turbo-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerGeneration", + "amount": 2.5, + "currency": "CNY", + "unit": "generation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "可灵 V2.5 Turbo Pro", + "description": "可灵高品质视频生成" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "1080p", + "4k" + ] + }, + "sourceRecord": "compute/providers/kling.json#kling-v2-5-turbo-pro", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-video.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-video_gen-provider-adapter", + "apiFormat": "kling-task-api", + "adapterId": "required-adapter:video_gen:kling" + } + }, + { + "apiModelId": "kling-v2", + "modelRef": "kling:kling-v2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "可灵 V2", + "description": "可灵文生视频旗舰模型" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "720p", + "1080p" + ] + }, + "sourceRecord": "compute/providers/kling.json#kling-v2", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-video.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-video_gen-provider-adapter", + "apiFormat": "kling-task-api", + "adapterId": "required-adapter:video_gen:kling" + } + }, + { + "apiModelId": "kling-v2-master", + "modelRef": "kling:kling-v2-master", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "可灵 V2 Master", + "description": "可灵最高品质视频生成" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "1080p", + "4k" + ] + }, + "sourceRecord": "compute/providers/kling.json#kling-v2-master", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-video.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-video_gen-provider-adapter", + "apiFormat": "kling-task-api", + "adapterId": "required-adapter:video_gen:kling" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/kling.json" + } + }, + { + "id": "provider-local-whisper", + "category": "providers", + "connection": { + "id": "provider-local-whisper", + "provider": "local-whisper", + "label": "本地 Whisper", + "baseUrl": "http://localhost:8080", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "asr" + ] + }, + "models": [ + { + "apiModelId": "whisper-large-v3", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "本地开源安装与云whisper-1不是同一合同；保留本地安装识别，不继承云端接口/额度。" + }, + "overrides": { + "serviceType": { + "value": [ + "asr" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/local-whisper.json#whisper-large-v3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "asr", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/local-whisper.json#whisper-large-v3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Whisper Large V3", + "description": "本地部署的 Whisper 大模型" + }, + "recordedExtra": { + "maxAudioLength": 7200, + "supportedInputFormats": [ + "wav", + "mp3", + "flac", + "ogg", + "webm" + ], + "streamingSupported": true + }, + "sourceRecord": "compute/providers/local-whisper.json#whisper-large-v3", + "automaticRouting": false, + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/asr-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:asr:local-whisper" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/local-whisper.json" + } + }, + { + "id": "provider-minimax-001", + "category": "providers", + "connection": { + "id": "provider-minimax-001", + "provider": "minimax", + "label": "MiniMax", + "baseUrl": "https://api.minimaxi.com/anthropic", + "apiFormat": "anthropic-messages", + "mediaBaseUrl": "https://api.minimaxi.com/v1", + "apiKeyRef": "minimax", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "fast", + "reasoning", + "vision", + "image_gen", + "tts", + "video_gen", + "music_gen" + ], + "tombstones": [ + "MiniMax-Text-01" + ] + }, + "models": [ + { + "apiModelId": "MiniMax-M3", + "modelRef": "minimax:MiniMax-M3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M3", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "vision", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M3", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 512000, + "inputPrice": 2.1, + "outputPrice": 8.4, + "cacheReadPrice": 0.42 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 4.2, + "outputPrice": 16.8, + "cacheReadPrice": 0.84 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.priorityPricingTiers", + "amount": [ + { + "maxInputTokens": 512000, + "inputPrice": 3.15, + "outputPrice": 12.6, + "cacheReadPrice": 0.63 + }, + { + "maxInputTokens": 1000000, + "inputPrice": 6.3, + "outputPrice": 25.2, + "cacheReadPrice": 1.26 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "MiniMax 国内开放平台按量计费永久五折价格。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M3", + "description": "MiniMax M3，前沿多模态 Coding 与 Agent 模型，1M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M3" + } + }, + { + "apiModelId": "MiniMax-M2.7", + "modelRef": "minimax:MiniMax-M2.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2.7", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.42, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7", + "description": "MiniMax 新一代旗舰模型，擅长复杂 Agent、软件工程与专业办公任务" + }, + "recordedExtra": { + "outputSpeedTps": 60 + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2.7" + } + }, + { + "apiModelId": "MiniMax-M2.7-highspeed", + "modelRef": "minimax:MiniMax-M2.7-highspeed", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2.7-highspeed", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.42, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7 高速版", + "description": "MiniMax M2.7 极速版，效果一致，输出更快，适合低延迟场景" + }, + "recordedExtra": { + "outputSpeedTps": 100 + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2.7-highspeed" + } + }, + { + "apiModelId": "MiniMax-M2.5", + "modelRef": "minimax:MiniMax-M2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.21, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5", + "description": "MiniMax M2.5，204.8K 上下文，支持深度推理和工具调用" + }, + "recordedExtra": { + "outputSpeedTps": 60 + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2.5" + } + }, + { + "apiModelId": "MiniMax-M2.5-highspeed", + "modelRef": "minimax:MiniMax-M2.5-highspeed", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2.5-highspeed", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.21, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5 高速版", + "description": "MiniMax M2.5 高速推理版本(100TPS)，适合低延迟场景" + }, + "recordedExtra": { + "outputSpeedTps": 100 + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2.5-highspeed" + } + }, + { + "apiModelId": "MiniMax-M2.1", + "modelRef": "minimax:MiniMax-M2.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2.1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.21, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.1", + "description": "MiniMax 上一代主力模型，200K 上下文，性价比优秀" + }, + "recordedExtra": { + "outputSpeedTps": 60 + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2.1" + } + }, + { + "apiModelId": "MiniMax-M2.1-highspeed", + "modelRef": "minimax:MiniMax-M2.1-highspeed", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2.1-highspeed", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.21, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.1 高速版", + "description": "MiniMax M2.1 高速推理版本，适合低延迟场景" + }, + "recordedExtra": { + "outputSpeedTps": 100 + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2.1-highspeed" + } + }, + { + "apiModelId": "MiniMax-M2", + "modelRef": "minimax:MiniMax-M2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#MiniMax-M2", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.21, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "legacy": true + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2", + "description": "MiniMax M2，优化编码和 Agent 工作流" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/minimax.json#MiniMax-M2" + } + }, + { + "apiModelId": "M2-her", + "modelRef": "minimax:M2-her", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/minimax.json#M2-her", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheReadPrice", + "amount": 0.21, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheWritePrice", + "amount": 2.625, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2 Her", + "description": "MiniMax 角色扮演对话模型，专为多轮角色对话设计" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/minimax.json#M2-her" + } + }, + { + "apiModelId": "image-01", + "modelRef": "minimax:image-01", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerImage", + "amount": 0.025, + "currency": "CNY", + "unit": "image", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Image 01", + "description": "MiniMax 文生图/图生图模型，支持主题参考（角色一致性），最长 1500 字符 prompt" + }, + "recordedExtra": { + "supportedAspectRatios": [ + "1:1", + "16:9", + "9:16", + "4:3", + "3:4", + "3:2", + "2:3", + "21:9" + ], + "maxPromptLength": 1500, + "maxBatchSize": 9, + "responseFormats": [ + "url", + "base64" + ], + "urlExpiry": "24h" + }, + "sourceRecord": "compute/providers/minimax.json#image-01" + } + }, + { + "apiModelId": "image-01-live", + "modelRef": "minimax:image-01-live", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Image 01 Live", + "description": "MiniMax 文生图模型（Live 版），支持画风控制（漫画/元气/中世纪/水彩）" + }, + "recordedExtra": { + "supportedAspectRatios": [ + "1:1", + "16:9", + "9:16", + "4:3", + "3:4", + "3:2", + "2:3" + ], + "supportedStyles": [ + "漫画", + "元气", + "中世纪", + "水彩" + ], + "maxPromptLength": 1500, + "maxBatchSize": 9, + "responseFormats": [ + "url", + "base64" + ], + "urlExpiry": "24h" + }, + "sourceRecord": "compute/providers/minimax.json#image-01-live" + } + }, + { + "apiModelId": "speech-2.8-hd", + "modelRef": "minimax:speech-2.8-hd", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerMillionCharacters", + "amount": 350, + "currency": "CNY", + "unit": "million-characters", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Speech 2.8 HD", + "description": "MiniMax 最新高清语音合成模型，支持 40+ 语言、100+ 预置音色，毫秒级延迟" + }, + "recordedExtra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ] + }, + "sourceRecord": "compute/providers/minimax.json#speech-2.8-hd", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "speech-2.8-turbo", + "modelRef": "minimax:speech-2.8-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerMillionCharacters", + "amount": 200, + "currency": "CNY", + "unit": "million-characters", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Speech 2.8 Turbo", + "description": "MiniMax 快速语音合成模型，速度优先，适合实时场景" + }, + "recordedExtra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ] + }, + "sourceRecord": "compute/providers/minimax.json#speech-2.8-turbo", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "speech-2.6-hd", + "modelRef": "minimax:speech-2.6-hd", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerMillionCharacters", + "amount": 350, + "currency": "CNY", + "unit": "million-characters", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Speech 2.6 HD", + "description": "MiniMax 高清语音合成模型（2.6 版），音质优秀，生成速度快" + }, + "recordedExtra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ] + }, + "sourceRecord": "compute/providers/minimax.json#speech-2.6-hd", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "speech-2.6-turbo", + "modelRef": "minimax:speech-2.6-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerMillionCharacters", + "amount": 200, + "currency": "CNY", + "unit": "million-characters", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Speech 2.6 Turbo", + "description": "MiniMax 快速语音合成模型（2.6 版），低延迟响应" + }, + "recordedExtra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ] + }, + "sourceRecord": "compute/providers/minimax.json#speech-2.6-turbo", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "speech-02-hd", + "modelRef": "minimax:speech-02-hd", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerMillionCharacters", + "amount": 350, + "currency": "CNY", + "unit": "million-characters", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "legacy": true + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Speech 02 HD", + "description": "MiniMax 高清语音合成模型（历史版本），韵律稳定性优秀" + }, + "recordedExtra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ] + }, + "sourceRecord": "compute/providers/minimax.json#speech-02-hd", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "speech-02-turbo", + "modelRef": "minimax:speech-02-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerMillionCharacters", + "amount": 200, + "currency": "CNY", + "unit": "million-characters", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "legacy": true + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Speech 02 Turbo", + "description": "MiniMax 快速语音合成模型（历史版本），增强小语种支持" + }, + "recordedExtra": { + "maxTextLength": 10000, + "supportedFormats": [ + "mp3", + "wav", + "pcm", + "flac", + "aac", + "ogg" + ] + }, + "sourceRecord": "compute/providers/minimax.json#speech-02-turbo", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "MiniMax-Hailuo-2.3", + "modelRef": "minimax:MiniMax-Hailuo-2.3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 2.3", + "description": "MiniMax 最新视频生成模型，支持文生视频和图生视频，支持 15 种运镜指令控制" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "768p", + "1080p" + ], + "maxPromptLength": 2000, + "asyncMode": true + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-Hailuo-2.3" + } + }, + { + "apiModelId": "MiniMax-Hailuo-2.3-Fast", + "modelRef": "minimax:MiniMax-Hailuo-2.3-Fast", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 2.3 快速版", + "description": "海螺视频快速版本（仅图生视频），生成速度更快" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "768p", + "1080p" + ], + "asyncMode": true + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-Hailuo-2.3-Fast" + } + }, + { + "apiModelId": "T2V-01-Director", + "modelRef": "minimax:T2V-01-Director", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown", + "legacy": true + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 Director", + "description": "导演级文生视频模型，精准运镜控制（历史版本）" + }, + "recordedExtra": { + "maxVideoDuration": 6, + "supportedResolutions": [ + "720p" + ], + "asyncMode": true + }, + "sourceRecord": "compute/providers/minimax.json#T2V-01-Director" + } + }, + { + "apiModelId": "MiniMax-Hailuo-02", + "modelRef": "minimax:MiniMax-Hailuo-02", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown", + "legacy": true + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 02", + "description": "海螺视频 02，原生 1080P 生成，指令跟随能力强（历史版本）" + }, + "recordedExtra": { + "maxVideoDuration": 10, + "supportedResolutions": [ + "512p", + "768p", + "1080p" + ], + "asyncMode": true + }, + "sourceRecord": "compute/providers/minimax.json#MiniMax-Hailuo-02" + } + }, + { + "apiModelId": "S2V-01", + "modelRef": "minimax:S2V-01", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 主体参考", + "description": "主体参考视频生成模型，基于参考图保持角色一致性" + }, + "recordedExtra": { + "asyncMode": true + }, + "sourceRecord": "compute/providers/minimax.json#S2V-01" + } + }, + { + "apiModelId": "music-2.6", + "modelRef": "minimax:music-2.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerSongUpToFiveMinutes", + "amount": 1, + "currency": "CNY", + "unit": "song-up-to-five-minutes", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Music 2.6", + "description": "MiniMax AI 音乐生成模型，支持音乐灵感描述和歌词输入，可生成带人声或纯音乐；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "recordedExtra": { + "maxPromptLength": 2000, + "maxLyricsLength": 3500, + "outputFormats": [ + "hex", + "url" + ], + "supportedLyricsStructures": [ + "[Intro]", + "[Verse]", + "[Pre Chorus]", + "[Chorus]", + "[Interlude]", + "[Bridge]", + "[Outro]", + "[Hook]", + "[Solo]", + "[Inst]" + ] + }, + "sourceRecord": "compute/providers/minimax.json#music-2.6" + } + }, + { + "apiModelId": "music-2.5+", + "modelRef": "minimax:music-2.5+", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerSongUpToFiveMinutes", + "amount": 1, + "currency": "CNY", + "unit": "song-up-to-five-minutes", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Music 2.5+", + "description": "MiniMax AI 音乐生成模型（2.5+ 版），音质提升；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "recordedExtra": { + "outputFormats": [ + "hex", + "url" + ] + }, + "sourceRecord": "compute/providers/minimax.json#music-2.5+" + } + }, + { + "apiModelId": "music-2.5", + "modelRef": "minimax:music-2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricePerSongUpToFiveMinutes", + "amount": 1, + "currency": "CNY", + "unit": "song-up-to-five-minutes", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Music 2.5", + "description": "MiniMax AI 音乐生成模型（2.5 版）；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "recordedExtra": { + "outputFormats": [ + "hex", + "url" + ] + }, + "sourceRecord": "compute/providers/minimax.json#music-2.5" + } + }, + { + "apiModelId": "music-cover", + "modelRef": "minimax:music-cover", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax Music Cover", + "description": "MiniMax AI 翻唱模型，基于参考音频生成翻唱版本，支持修改歌词；音乐付费 API 自2026-08-20起仅历史付费用户可继续使用，不向新用户开放；免费音乐接口已停止服务" + }, + "recordedExtra": { + "maxAudioDuration": 360, + "maxAudioSize": "50MB", + "supportedFormats": [ + "mp3", + "wav", + "flac" + ] + }, + "sourceRecord": "compute/providers/minimax.json#music-cover" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/minimax.json" + } + }, + { + "id": "provider-mistral-001", + "category": "providers", + "connection": { + "id": "provider-mistral-001", + "provider": "mistral", + "label": "Mistral AI", + "baseUrl": "https://api.mistral.ai/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "mistral", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "fast" + ] + }, + "models": [ + { + "apiModelId": "mistral-medium-3-5", + "modelRef": "mistral:mistral-medium-3.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 7.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Mistral Medium 3.5", + "description": "Mistral Medium 3.5，官方 API ID mistral-medium-3-5，256K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/mistral.json#mistral-medium-3-5" + } + }, + { + "apiModelId": "mistral-medium-latest", + "modelRef": "mistral:mistral-medium-3.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#mistral-medium-latest", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#mistral-medium-latest", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#mistral-medium-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 7.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Mistral Medium 3.5", + "description": "Mistral 前沿多模态模型，面向 Agentic 和 Coding 场景优化，256K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/mistral.json#mistral-medium-latest" + } + }, + { + "apiModelId": "mistral-large-latest", + "modelRef": "mistral:mistral-large-latest", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#mistral-large-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Mistral Large 3", + "description": "Mistral 旗舰模型，256K 上下文，支持视觉和工具调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/mistral.json#mistral-large-latest" + } + }, + { + "apiModelId": "mistral-small-latest", + "modelRef": "mistral:mistral-small-latest", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "code", + "vision", + "reasoning", + "fast", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#mistral-small-latest", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#mistral-small-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.15, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.6, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Mistral Small 4", + "description": "Mistral Small 4，指令、推理与编码混合模型，256K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/mistral.json#mistral-small-latest" + } + }, + { + "apiModelId": "codestral-latest", + "modelRef": "mistral:codestral-latest", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/mistral.json#codestral-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.9, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Codestral", + "description": "Mistral 专业代码模型，支持 Fill-in-the-Middle 补全" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/mistral.json#codestral-latest" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/mistral.json" + } + }, + { + "id": "provider-moonshot-001", + "category": "providers", + "connection": { + "id": "provider-moonshot-001", + "provider": "moonshot", + "label": "月之暗面 Kimi", + "baseUrl": "https://api.moonshot.cn/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "moonshot", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning", + "vision" + ], + "tombstones": [ + "kimi-k2", + "kimi-k2-0905-preview", + "kimi-k2-0711-preview", + "kimi-k2-turbo-preview", + "kimi-k2-thinking", + "kimi-k2-thinking-turbo", + "kimi-latest", + "kimi-thinking-preview" + ] + }, + "models": [ + { + "apiModelId": "kimi-k3", + "modelRef": "moonshot:kimi-k3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "constraints": { + "thinkingOnly": true, + "samplingParametersDeprecated": true + }, + "thinking": { + "default": "enabled", + "disableSupported": false, + "preserveReasoningContent": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 20, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 100, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "CNY per 1M tokens: input 20, output 100, cache hit 2. Cache writes: TTL 5min 20, TTL 1h 40; recorded separately from cache-hit billing.", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K3", + "description": "Kimi K3，1M 上下文，始终开启思考，low/high/max 三档（默认 max）；输出上限与输入共享窗口，默认 max_completion_tokens=131072" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/moonshot.json#kimi-k3" + } + }, + { + "apiModelId": "kimi-k2.7-code", + "modelRef": "moonshot:kimi-k2.7-code", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 32768, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled", + "disableSupported": false, + "preserveReasoningContent": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 6.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 27, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1.3, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.7 Code", + "description": "月之暗面 Kimi K2.7 Code，当前最强 Coding 模型，256K 上下文，支持文本/图片/视频输入，始终启用思考模式" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#kimi-k2.7-code" + } + }, + { + "apiModelId": "kimi-k2.7-code-highspeed", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 32768, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled", + "disableSupported": false, + "preserveReasoningContent": true + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 13, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 54, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 2.6, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.7 Code HighSpeed", + "description": "月之暗面 Kimi K2.7 Code 高速版，与 K2.7 Code 参数一致，输出速度约 180 Tokens/s，短上下文可达 260 Tokens/s" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#kimi-k2.7-code-highspeed", + "automaticRouting": false + } + }, + { + "apiModelId": "kimi-k2.6", + "modelRef": "moonshot:kimi-k2.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.6", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 32768, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.6", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.6", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.6", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled", + "disabledTemperature": 0.6 + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 6.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 27, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1.1, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.6", + "description": "月之暗面 Kimi K2.6，多模态模型，长程代码编写与 Agent 自主执行能力增强，256K 上下文，支持思考/非思考模式" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#kimi-k2.6" + } + }, + { + "apiModelId": "kimi-k2.5", + "modelRef": "moonshot:kimi-k2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.5", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "tool_use", + "agent", + "long_context", + "vision", + "video_understanding", + "image_understanding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.5", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#kimi-k2.5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled", + "disabledTemperature": 0.6 + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 21, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.7, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.5", + "description": "月之暗面 Kimi K2.5，原生多模态模型，256K 上下文，支持思考/非思考模式" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ], + "supportedVideoFormats": [ + "mp4", + "mpeg", + "mov", + "avi", + "x-flv", + "mpg", + "webm", + "wmv", + "3gpp" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#kimi-k2.5" + } + }, + { + "apiModelId": "moonshot-v1-8k", + "modelRef": "moonshot:moonshot-v1-8k", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Moonshot V1 8K", + "description": "月之暗面 Moonshot V1 生成模型，8K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/moonshot.json#moonshot-v1-8k" + } + }, + { + "apiModelId": "moonshot-v1-32k", + "modelRef": "moonshot:moonshot-v1-32k", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Moonshot V1 32K", + "description": "月之暗面 Moonshot V1 生成模型，32K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/moonshot.json#moonshot-v1-32k" + } + }, + { + "apiModelId": "moonshot-v1-128k", + "modelRef": "moonshot:moonshot-v1-128k", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Moonshot V1 128K", + "description": "月之暗面 Moonshot V1 生成模型，128K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/moonshot.json#moonshot-v1-128k" + } + }, + { + "apiModelId": "moonshot-v1-8k-vision-preview", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "视觉变体不能继承文本8K family；无已证精确绑定时保留接入局部窗口和模态。" + }, + "overrides": { + "serviceType": { + "value": [ + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 4096, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "image_understanding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 0, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Moonshot V1 8K Vision Preview", + "description": "月之暗面 Moonshot V1 视觉模型，理解图片内容并输出文本，8K 上下文" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#moonshot-v1-8k-vision-preview", + "automaticRouting": false + } + }, + { + "apiModelId": "moonshot-v1-32k-vision-preview", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "视觉变体不能继承文本8K family；无已证精确绑定时保留接入局部窗口和模态。" + }, + "overrides": { + "serviceType": { + "value": [ + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 32768, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 4096, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "image_understanding", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 0, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Moonshot V1 32K Vision Preview", + "description": "月之暗面 Moonshot V1 视觉模型，理解图片内容并输出文本，32K 上下文" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#moonshot-v1-32k-vision-preview", + "automaticRouting": false + } + }, + { + "apiModelId": "moonshot-v1-128k-vision-preview", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "视觉变体不能继承文本8K family；无已证精确绑定时保留接入局部窗口和模态。" + }, + "overrides": { + "serviceType": { + "value": [ + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 4096, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "image_understanding", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 0, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Moonshot V1 128K Vision Preview", + "description": "月之暗面 Moonshot V1 视觉模型，理解图片内容并输出文本，128K 上下文" + }, + "recordedExtra": { + "supportedImageFormats": [ + "png", + "jpeg", + "webp", + "gif" + ] + }, + "sourceRecord": "compute/providers/moonshot.json#moonshot-v1-128k-vision-preview", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/providers/moonshot.json" + } + }, + { + "id": "provider-local-ollama", + "category": "providers", + "connection": { + "id": "provider-local-ollama", + "provider": "ollama", + "label": "本地 Ollama", + "baseUrl": "http://localhost:11434/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat" + ] + }, + "models": [ + { + "apiModelId": "llama3.1:70b", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/ollama.json#llama3.1:70b", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/ollama.json#llama3.1:70b", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/providers/ollama.json#llama3.1:70b", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "code", + "reasoning", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://ollama.com/library/llama3.1:70b", + "verification": "tool-use-officially-verified-other-tags-inherited" + }, + "defaultTemperature": { + "value": 0.8, + "reason": "preserved-access-contract", + "evidence": "compute/providers/ollama.json#llama3.1:70b", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.9, + "reason": "preserved-access-contract", + "evidence": "compute/providers/ollama.json#llama3.1:70b", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Llama 3.1 70B", + "description": "本地运行的 Llama 3.1 70B" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/ollama.json#llama3.1:70b", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/providers/ollama.json" + } + }, + { + "id": "provider-openai-codex-plan-001", + "category": "providers", + "connection": { + "id": "provider-openai-codex-plan-001", + "provider": "openai-codex", + "brandGroup": "openai", + "label": "ChatGPT 订阅 (Codex)", + "baseUrl": "https://chatgpt.com/backend-api", + "apiFormat": "openai-codex-responses", + "apiKeyRef": "openai-codex-access-token", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "accessMode": "coding-plan", + "credentialSource": "codex-cli", + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://chatgpt.com/codex/settings/usage" + } + }, + "services": [ + "chat", + "reasoning" + ], + "tombstones": [ + "gpt-5.3-codex", + "gpt-5.2-codex", + "gpt-5.1-codex-mini" + ] + }, + "models": [ + { + "apiModelId": "gpt-6.1-sol", + "modelRef": "openai:gpt-6.1-sol", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-6.1-sol", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6.1 Sol", + "description": "GPT-6.1 Sol，面向复杂编程、电脑使用和专业工作，1.05M 上下文；工具调用使用 Responses API", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-6.1-sol" + } + }, + { + "apiModelId": "gpt-6-sol", + "modelRef": "openai:gpt-6-sol", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-6-sol", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6 Sol", + "description": "ChatGPT 订阅额度计费，兼顾复杂编程、智能体工作流与成本的 GPT-6 模型；1.05M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-6-sol" + } + }, + { + "apiModelId": "gpt-6-luna", + "modelRef": "openai:gpt-6-luna", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-6-luna", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6 Luna", + "description": "ChatGPT 订阅额度计费，面向明确任务和高吞吐场景的 GPT-6 轻量模型；1.05M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-6-luna" + } + }, + { + "apiModelId": "gpt-6-astra", + "modelRef": "openai:gpt-6-astra", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6 Astra", + "description": "ChatGPT 订阅额度计费，面向复杂推理、编程和长程智能体任务，1.05M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-6-astra" + } + }, + { + "apiModelId": "gpt-reserve", + "modelRef": "openai:gpt-reserve", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-reserve", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-Reserve", + "description": "ChatGPT 订阅额度计费，参数暂按 GPT-5.6 Luna 配置，待官方规格确认，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-reserve" + } + }, + { + "apiModelId": "gpt-5.6-sol", + "modelRef": "openai:gpt-5.6-sol", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.6-sol", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "low" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.6 Sol", + "description": "ChatGPT 订阅额度计费，新一代旗舰模型，复杂编程/计算机操作/网络安全能力最强，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.6-sol" + } + }, + { + "apiModelId": "gpt-5.6-terra", + "modelRef": "openai:gpt-5.6-terra", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.6-terra", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.6 Terra", + "description": "ChatGPT 订阅额度计费，均衡型模型，日常工作性能对齐 GPT-5.5 且成本更低，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.6-terra" + } + }, + { + "apiModelId": "gpt-5.6-luna", + "modelRef": "openai:gpt-5.6-luna", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.6-luna", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.6 Luna", + "description": "ChatGPT 订阅额度计费，GPT-5.6 家族中最快最省成本的档位，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.6-luna" + } + }, + { + "apiModelId": "gpt-5.5", + "modelRef": "openai:gpt-5.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.5", + "description": "ChatGPT 订阅额度计费，上一代旗舰模型，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.5" + } + }, + { + "apiModelId": "gpt-5.4", + "modelRef": "openai:gpt-5.4", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.4", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.4", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.4", + "description": "ChatGPT 订阅额度计费，高阶通用模型，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.4" + } + }, + { + "apiModelId": "gpt-5.4-mini", + "modelRef": "openai:gpt-5.4-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 272000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai-codex.json#gpt-5.4-mini", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.4 Mini", + "description": "ChatGPT 订阅额度计费，轻量高性价比模型，是 gpt-5.1-codex-mini 的官方迁移目标，272K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.4-mini" + } + }, + { + "apiModelId": "gpt-5.3-codex-spark", + "modelRef": "openai:gpt-5.3-codex-spark", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.3 Codex Spark（Pro 专属）", + "description": "ChatGPT Pro 订阅专属研究预览模型，面向实时编程迭代优化，>1000 tok/s，纯文本，128K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai-codex.json#gpt-5.3-codex-spark" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/openai-codex.json" + } + }, + { + "id": "provider-openai-001", + "category": "providers", + "connection": { + "id": "provider-openai-001", + "provider": "openai", + "label": "OpenAI GPT", + "baseUrl": "https://api.openai.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "openai", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "fast", + "embedding", + "tts", + "asr", + "vision", + "responses", + "image_gen", + "omni", + "realtime_voice", + "simultaneous_interpret", + "translation" + ], + "tombstones": [ + "dall-e-3", + "gpt-4o-realtime", + "gpt-4o-realtime-preview" + ] + }, + "models": [ + { + "apiModelId": "gpt-6.1-sol", + "modelRef": "openai:gpt-6.1-sol", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6.1-sol", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6.1-sol", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 2, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + }, + "responsesOnly": true + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.1, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Cache writes $2.50/M. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request. Fast mode is 2x; Batch and Flex are 50% lower. Tool calling requires Responses API; none/minimal reasoning are unsupported.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6.1 Sol", + "description": "GPT-6.1 Sol，面向复杂编程、电脑使用和专业工作，1.05M 上下文；工具调用使用 Responses API" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-6.1-sol" + } + }, + { + "apiModelId": "gpt-image-2.5-sunburst", + "modelRef": "openai:gpt-image-2.5-sunburst", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageInputPrice", + "amount": 8, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.textCachedInputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageCachedInputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Text input $5/M tokens, image input $8/M tokens, image output $30/M tokens; quality supports low, medium, high, xhigh, max, auto.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT Image 2.5 Sunburst", + "description": "图像生成与编辑模型，适合对编辑精度要求高的任务" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-image-2.5-sunburst", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:openai" + } + }, + { + "apiModelId": "gpt-image-2.5-flare", + "modelRef": "openai:gpt-image-2.5-flare", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageInputPrice", + "amount": 8, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.textCachedInputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageCachedInputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Text input $5/M tokens, image input $8/M tokens, image output $30/M tokens; quality supports low, medium, high, xhigh, max, auto.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT Image 2.5 Flare", + "description": "快速图像生成与编辑模型，适合日常高质量出图" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-image-2.5-flare", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:openai" + } + }, + { + "apiModelId": "gpt-6-sol", + "modelRef": "openai:gpt-6-sol", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6-sol", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6-sol", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 2, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request. Chat Completions function calling requires reasoning_effort=none.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6 Sol", + "description": "兼顾复杂编程、智能体工作流与成本的 GPT-6 模型；1.05M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-6-sol" + } + }, + { + "apiModelId": "gpt-6-luna", + "modelRef": "openai:gpt-6-luna", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "fast", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6-luna", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6-luna", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 3, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.01, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request. Chat Completions function calling requires reasoning_effort=none.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6 Luna", + "description": "面向明确任务和高吞吐场景的 GPT-6 轻量模型；1.05M 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-6-luna" + } + }, + { + "apiModelId": "gpt-6-astra", + "modelRef": "openai:gpt-6-astra", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6-astra", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-6-astra", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 1, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "medium" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 50, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 1, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens. Cache writes cost $12.50. Above 272K input tokens, input and cache rates are 2x and output rates are 1.5x for the full request.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-6 Astra", + "description": "OpenAI GPT-6 Astra，面向复杂推理、编程和长程智能体任务，1.05M 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-6-astra" + } + }, + { + "apiModelId": "gpt-5.5", + "modelRef": "openai:gpt-5.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 1, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ], + "defaultEffort": "medium" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens; long-context requests may use higher-tier pricing per OpenAI pricing page.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.5", + "description": "OpenAI GPT-5.5 前沿模型，面向复杂推理、编码和专业工作流，1M 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.5" + } + }, + { + "apiModelId": "gpt-5.5-pro", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.5-pro", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1050000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.5-pro", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 128000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.5-pro", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.5-pro", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 2, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "medium", + "high", + "xhigh" + ], + "defaultEffort": "high" + }, + "responsesOnly": true + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 180, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "No cached-input discount is listed for GPT-5.5 Pro on the official pricing page.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.5 Pro", + "description": "OpenAI GPT-5.5 Pro，最高能力推理模型，仅面向 Responses API" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.5-pro", + "automaticRouting": false + } + }, + { + "apiModelId": "gpt-5.4", + "modelRef": "openai:gpt-5.4", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.4", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.4", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 3, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ] + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 15, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.25, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens; long-context requests may use higher-tier pricing per OpenAI pricing page.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.4", + "description": "OpenAI GPT-5.4，高性价比前沿模型，1M 上下文，支持文本和图像输入" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.4" + } + }, + { + "apiModelId": "gpt-5.4-pro", + "modelRef": "openai:gpt-5.4-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.4-pro", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "ultra_long_context", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.4-pro", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 4, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "medium", + "high", + "xhigh" + ], + "defaultEffort": "medium" + }, + "responsesOnly": true + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 180, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.4 Pro", + "description": "OpenAI GPT-5.4 Pro，专业推理与长程任务模型，仅面向 Responses API" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.4-pro" + } + }, + { + "apiModelId": "gpt-5.4-mini", + "modelRef": "openai:gpt-5.4-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "fast", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.4-mini", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 5, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ] + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 4.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.075, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.4 Mini", + "description": "OpenAI GPT-5.4 Mini，低延迟高性价比模型，400K 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.4-mini" + } + }, + { + "apiModelId": "gpt-5.4-nano", + "modelRef": "openai:gpt-5.4-nano", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "fast", + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-5.4-nano", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "fast", + "long_context", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://developers.openai.com/api/docs/models/gpt-5.4-nano.md", + "verification": "tool-use-officially-verified-other-tags-inherited" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 6, + "toolType": "web_search" + }, + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high", + "xhigh" + ] + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.02, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.4 Nano", + "description": "OpenAI GPT-5.4 Nano，最低成本极速模型，400K 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.4-nano" + } + }, + { + "apiModelId": "gpt-image-2", + "modelRef": "openai:gpt-image-2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "image_generation", + "image_editing", + "text_rendering", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-image-2", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageInputPrice", + "amount": 8, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.textCachedInputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageCachedInputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Text input is $5/M tokens, image input is $8/M tokens, image output is $30/M tokens.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT Image 2", + "description": "OpenAI GPT Image 2 图像生成和编辑模型，支持文本输入、图像输入与图像输出" + }, + "recordedExtra": { + "snapshot": "gpt-image-2-2026-04-21" + }, + "sourceRecord": "compute/providers/openai.json#gpt-image-2", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:openai" + } + }, + { + "apiModelId": "gpt-realtime-2.1", + "modelRef": "openai:gpt-realtime-2.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "omni", + "realtime_voice" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-realtime-2.1", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tts", + "asr", + "realtime", + "speech_to_speech", + "reasoning", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-realtime-2.1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 24, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.textCachedInputPrice", + "amount": 0.4, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageInputPrice", + "amount": 5, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageCachedInputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioInputPrice", + "amount": 32, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioCachedInputPrice", + "amount": 0.4, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioOutputPrice", + "amount": 64, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT Realtime 2.1", + "description": "OpenAI 最新实时多模态语音模型，支持文本、图像和音频输入输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-realtime-2.1", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/speech-core/src/providers/index.ts", + "sha256": "f57b8fc13ac7465a5aca62610d555adbbc0078643a335568484e38f1610c0e1d", + "assertion": "createBuiltinSpeechAdapters只注册ASR/TTS；OpenAI均batch endpoint；无Realtime WebSocket/translation/session适配器。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "兼容表OpenAI仅精确tts-1/tts-1-hd、whisper-1，Realtime3型号没有语音兼容profile。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile无显式有效speech身份时只调用兼容表；未命中fail missing_model_profile，不能以Provider.apiFormat泛化支持Realtime。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-model-specific-realtime-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:omni,realtime_voice:openai" + } + }, + { + "apiModelId": "gpt-realtime-2.1-mini", + "modelRef": "openai:gpt-realtime-2.1-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "realtime_voice" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-realtime-2.1-mini", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tts", + "asr", + "realtime", + "speech_to_speech", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-realtime-2.1-mini", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.6, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.textCachedInputPrice", + "amount": 0.06, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageInputPrice", + "amount": 0.8, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.imageCachedInputPrice", + "amount": 0.08, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioInputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioCachedInputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.audioOutputPrice", + "amount": 20, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT Realtime 2.1 Mini", + "description": "OpenAI 低成本实时语音模型，适合实时语音交互" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-realtime-2.1-mini", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/speech-core/src/providers/index.ts", + "sha256": "f57b8fc13ac7465a5aca62610d555adbbc0078643a335568484e38f1610c0e1d", + "assertion": "createBuiltinSpeechAdapters只注册ASR/TTS；OpenAI均batch endpoint；无Realtime WebSocket/translation/session适配器。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "兼容表OpenAI仅精确tts-1/tts-1-hd、whisper-1，Realtime3型号没有语音兼容profile。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile无显式有效speech身份时只调用兼容表；未命中fail missing_model_profile，不能以Provider.apiFormat泛化支持Realtime。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-model-specific-realtime-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:realtime_voice:openai" + } + }, + { + "apiModelId": "gpt-realtime-translate", + "modelRef": "openai:gpt-realtime-translate", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "simultaneous_interpret", + "translation" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-realtime-translate", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "translation", + "simultaneous_interpretation", + "asr", + "realtime" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-realtime-translate", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "extra.audioOutputPricePerMinute", + "amount": 0.034, + "currency": "USD", + "unit": "minute", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "OpenAI lists this model at $0.034 per minute.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT Realtime Translate", + "description": "OpenAI 实时语音翻译模型，按分钟计费" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-realtime-translate", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/speech-core/src/providers/index.ts", + "sha256": "f57b8fc13ac7465a5aca62610d555adbbc0078643a335568484e38f1610c0e1d", + "assertion": "createBuiltinSpeechAdapters只注册ASR/TTS；OpenAI均batch endpoint；无Realtime WebSocket/translation/session适配器。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "兼容表OpenAI仅精确tts-1/tts-1-hd、whisper-1，Realtime3型号没有语音兼容profile。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile无显式有效speech身份时只调用兼容表；未命中fail missing_model_profile，不能以Provider.apiFormat泛化支持Realtime。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-model-specific-realtime-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:simultaneous_interpret,translation:openai" + } + }, + { + "apiModelId": "gpt-5.2", + "modelRef": "openai:gpt-5.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 14, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.2", + "description": "OpenAI GPT-5.2，400K 上下文，面向编码和智能体任务" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.2" + } + }, + { + "apiModelId": "gpt-5.2-pro", + "modelRef": "openai:gpt-5.2-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 21, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 168, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.2 Pro", + "description": "OpenAI GPT-5.2 Pro 专业版，最高性能" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.2-pro" + } + }, + { + "apiModelId": "gpt-5.1", + "modelRef": "openai:gpt-5.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5.1", + "description": "OpenAI GPT-5.1，400K 上下文，高性能旗舰" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5.1" + } + }, + { + "apiModelId": "gpt-5", + "modelRef": "openai:gpt-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 7, + "toolType": "web_search" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5", + "description": "OpenAI GPT-5 新一代旗舰模型，统一所有模型能力" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5" + } + }, + { + "apiModelId": "gpt-5-pro", + "modelRef": "openai:gpt-5-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 15, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 120, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5 Pro", + "description": "OpenAI GPT-5 Pro 专业版高性能推理模型" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5-pro" + } + }, + { + "apiModelId": "gpt-5-mini", + "modelRef": "openai:gpt-5-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5-mini", + "description": "高性价比模型，适合日常对话，400K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5-mini" + } + }, + { + "apiModelId": "gpt-5-nano", + "modelRef": "openai:gpt-5-nano", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.05, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-5-nano", + "description": "极速响应模型，适合简单任务，400K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-5-nano" + } + }, + { + "apiModelId": "gpt-4.1", + "modelRef": "openai:gpt-4.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-4.1", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 8, + "toolType": "web_search" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-4.1", + "description": "OpenAI GPT-4.1，支持 1M 上下文，编码和指令遵循能力大幅提升" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-4.1" + } + }, + { + "apiModelId": "gpt-4.1-mini", + "modelRef": "openai:gpt-4.1-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-4.1-mini", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 9, + "toolType": "web_search" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.6, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-4.1 mini", + "description": "OpenAI GPT-4.1 mini，1M 上下文，高性价比" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-4.1-mini" + } + }, + { + "apiModelId": "gpt-4.1-nano", + "modelRef": "openai:gpt-4.1-nano", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-4.1-nano", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-4.1 nano", + "description": "OpenAI GPT-4.1 nano，1M 上下文，极致性价比" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-4.1-nano" + } + }, + { + "apiModelId": "gpt-4o", + "modelRef": "openai:gpt-4o", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-4o", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-4o", + "description": "多模态模型，支持图像理解" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-4o" + } + }, + { + "apiModelId": "gpt-4o-mini", + "modelRef": "openai:gpt-4o-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#gpt-4o-mini", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.15, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.6, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GPT-4o mini", + "description": "OpenAI GPT-4o mini，高性价比多模态模型" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#gpt-4o-mini" + } + }, + { + "apiModelId": "text-embedding-3-small", + "modelRef": "openai:text-embedding-3-small", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.02, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "text-embedding-3-small", + "description": "轻量文本向量化模型，1536 维" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#text-embedding-3-small" + } + }, + { + "apiModelId": "text-embedding-3-large", + "modelRef": "openai:text-embedding-3-large", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.13, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "text-embedding-3-large", + "description": "高精度文本向量化模型，3072 维" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#text-embedding-3-large" + } + }, + { + "apiModelId": "tts-1", + "modelRef": "openai:tts-1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 15, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "TTS-1", + "description": "标准语音合成" + }, + "recordedExtra": { + "voices": [ + "alloy", + "echo", + "fable", + "onyx", + "nova", + "shimmer" + ], + "supportedAudioFormats": [ + "mp3", + "opus", + "aac", + "flac" + ] + }, + "sourceRecord": "compute/providers/openai.json#tts-1" + } + }, + { + "apiModelId": "tts-1-hd", + "modelRef": "openai:tts-1-hd", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 30, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "TTS-1-HD", + "description": "高清语音合成" + }, + "recordedExtra": { + "voices": [ + "alloy", + "echo", + "fable", + "onyx", + "nova", + "shimmer" + ], + "supportedAudioFormats": [ + "mp3", + "opus", + "aac", + "flac" + ] + }, + "sourceRecord": "compute/providers/openai.json#tts-1-hd" + } + }, + { + "apiModelId": "whisper-1", + "modelRef": "openai:whisper-1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.006, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Whisper", + "description": "通用语音识别" + }, + "recordedExtra": { + "maxAudioLength": 600, + "supportedInputFormats": [ + "mp3", + "mp4", + "mpeg", + "mpga", + "m4a", + "wav", + "webm" + ], + "streamingSupported": false + }, + "sourceRecord": "compute/providers/openai.json#whisper-1" + } + }, + { + "apiModelId": "o3", + "modelRef": "openai:o3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#o3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "o3", + "description": "OpenAI o3 推理模型，200K 上下文，支持工具调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#o3" + } + }, + { + "apiModelId": "o3-pro", + "modelRef": "openai:o3-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#o3-pro", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 20, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 80, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "o3-pro", + "description": "OpenAI o3-pro 高级推理模型，200K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#o3-pro" + } + }, + { + "apiModelId": "o3-mini", + "modelRef": "openai:o3-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#o3-mini", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 4.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "o3-mini", + "description": "OpenAI o3-mini 推理模型，200K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#o3-mini" + } + }, + { + "apiModelId": "o4-mini", + "modelRef": "openai:o4-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "responses" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openai.json#o4-mini", + "verification": "inherited-not-reverified" + } + }, + "request": { + "serverSideWebSearch": { + "enabled": true, + "dialect": "openai-responses", + "fallbackPriority": 10, + "toolType": "web_search" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 4.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.serverSideWebSearch.searchRequestPriceUsd", + "amount": 0.01, + "currency": "USD", + "unit": "search-request", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "o4-mini", + "description": "OpenAI o4-mini 推理模型，200K 上下文，支持工具调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openai.json#o4-mini" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/openai.json" + } + }, + { + "id": "provider-openrouter-001", + "category": "providers", + "connection": { + "id": "provider-openrouter-001", + "provider": "openrouter", + "label": "OpenRouter", + "baseUrl": "https://openrouter.ai/api/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "openrouter", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "reasoning", + "vision" + ], + "tombstones": [ + "stealth/ox-alpha", + "openai/gpt-oss-120b:free", + "qwen/qwen3-coder:free" + ] + }, + "models": [ + { + "apiModelId": "cohere/command-a-plus", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "聚合平台请求ID有独立托管合同；没有该精确ID已证alias时保留接入本地数值/能力，不通过剥前缀绑定原厂。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#cohere/command-a-plus", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 192000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#cohere/command-a-plus", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#cohere/command-a-plus", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 64000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#cohere/command-a-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Cohere: Command A+", + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#cohere/command-a-plus", + "automaticRouting": false + } + }, + { + "apiModelId": "mistralai/mistral-medium-3-5", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "聚合平台请求ID有独立托管合同；没有该精确ID已证alias时保留接入本地数值/能力，不通过剥前缀绑定原厂。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#mistralai/mistral-medium-3-5", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#mistralai/mistral-medium-3-5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#mistralai/mistral-medium-3-5", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 209715, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#mistralai/mistral-medium-3-5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 7.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Mistral: Mistral Medium 3.5", + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#mistralai/mistral-medium-3-5", + "automaticRouting": false + } + }, + { + "apiModelId": "google/gemini-3.8-flash", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "聚合平台请求ID有独立托管合同；没有该精确ID已证alias时保留接入本地数值/能力，不通过剥前缀绑定原厂。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#google/gemini-3.8-flash", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1048576, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#google/gemini-3.8-flash", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#google/gemini-3.8-flash", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 65536, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#google/gemini-3.8-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3.75, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Google: Gemini 3.8 Flash", + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#google/gemini-3.8-flash", + "automaticRouting": false + } + }, + { + "apiModelId": "minimax/minimax-m3", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "聚合平台请求ID有独立托管合同；没有该精确ID已证alias时保留接入本地数值/能力，不通过剥前缀绑定原厂。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#minimax/minimax-m3", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1048576, + "reason": "本次OpenRouter官方API原样返回独立context_length和top_provider.max_completion_tokens；接入值保留，不复制到原厂规格。顶层目录窗口与top_provider路由窗口不同，二者不得合并。", + "evidence": "https://openrouter.ai/api/v1/models", + "verification": "officially-verified-access-contract" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#minimax/minimax-m3", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 512000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#minimax/minimax-m3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.2, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax: MiniMax M3", + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#minimax/minimax-m3", + "automaticRouting": false + } + }, + { + "apiModelId": "qwen/qwen3.8-omni-flash", + "modelRef": "qwen:qwen3.8-omni-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#qwen/qwen3.8-omni-flash", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#qwen/qwen3.8-omni-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.15, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.47, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen: Qwen3.8 Omni Flash", + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#qwen/qwen3.8-omni-flash" + } + }, + { + "apiModelId": "qwen/qwen3.8-max-prime", + "modelRef": "qwen:qwen3.8-max-prime", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#qwen/qwen3.8-max-prime", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#qwen/qwen3.8-max-prime", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 12, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen: Qwen3.8 Max Prime", + "description": "OpenRouter 官方公共目录模型，接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#qwen/qwen3.8-max-prime" + } + }, + { + "apiModelId": "deepseek/deepseek-v4.1-flash", + "modelRef": "deepseek:deepseek-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#deepseek/deepseek-v4.1-flash", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1048576, + "reason": "本次OpenRouter官方API原样返回独立context_length和top_provider.max_completion_tokens；接入值保留，不复制到原厂规格。顶层目录窗口与top_provider路由窗口不同，二者不得合并。", + "evidence": "https://openrouter.ai/api/v1/models", + "verification": "officially-verified-access-contract" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#deepseek/deepseek-v4.1-flash", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 943718, + "reason": "本次OpenRouter官方API原样返回独立context_length和top_provider.max_completion_tokens；接入值保留，不复制到原厂规格。顶层目录窗口与top_provider路由窗口不同，二者不得合并。", + "evidence": "https://openrouter.ai/api/v1/models", + "verification": "officially-verified-access-contract" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.2, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek: DeepSeek V4.1 Flash", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#deepseek/deepseek-v4.1-flash" + } + }, + { + "apiModelId": "tencent/hy4-preview", + "modelRef": "tencent:hy4-preview", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#tencent/hy4-preview", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 64000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#tencent/hy4-preview", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.834, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.501, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Tencent: Hy4 preview", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#tencent/hy4-preview" + } + }, + { + "apiModelId": "z-ai/glm-5.3-flash", + "modelRef": "zhipu:glm-5.3-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#z-ai/glm-5.3-flash", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#z-ai/glm-5.3-flash", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 943717, + "reason": "本次OpenRouter官方API原样返回独立context_length和top_provider.max_completion_tokens；接入值保留，不复制到原厂规格。顶层目录窗口与top_provider路由窗口不同，二者不得合并。", + "evidence": "https://openrouter.ai/api/v1/models", + "verification": "officially-verified-access-contract" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.15, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Z.ai: GLM 5.3 Flash", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#z-ai/glm-5.3-flash" + } + }, + { + "apiModelId": "z-ai/glm-5.3", + "modelRef": "zhipu:glm-5.3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#z-ai/glm-5.3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 4.4, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Z.ai: GLM 5.3", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#z-ai/glm-5.3" + } + }, + { + "apiModelId": "moonshotai/kimi-k3", + "modelRef": "moonshot:kimi-k3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#moonshotai/kimi-k3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#moonshotai/kimi-k3", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 943718, + "reason": "本次OpenRouter官方API原样返回独立context_length和top_provider.max_completion_tokens；接入值保留，不复制到原厂规格。顶层目录窗口与top_provider路由窗口不同，二者不得合并。", + "evidence": "https://openrouter.ai/api/v1/models", + "verification": "officially-verified-access-contract" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.7, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 13.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MoonshotAI: Kimi K3", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#moonshotai/kimi-k3" + } + }, + { + "apiModelId": "anthropic/claude-sonnet-5.5", + "modelRef": "anthropic:claude-sonnet-5-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#anthropic/claude-sonnet-5.5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Anthropic: Claude Sonnet 5.5", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#anthropic/claude-sonnet-5.5" + } + }, + { + "apiModelId": "anthropic/claude-opus-5.5", + "modelRef": "anthropic:claude-opus-5-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#anthropic/claude-opus-5.5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Anthropic: Claude Opus 5.5", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#anthropic/claude-opus-5.5" + } + }, + { + "apiModelId": "openai/gpt-6.1-sol", + "modelRef": "openai:gpt-6.1-sol", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "vision", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#openai/gpt-6.1-sol", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "USD per 1M tokens; OpenRouter access-specific pricing.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "OpenAI: GPT-6.1 Sol", + "description": "OpenRouter 官方公共目录模型；接入参数与计价以该平台为准" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#openai/gpt-6.1-sol" + } + }, + { + "apiModelId": "openrouter/auto", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#openrouter/auto", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 2000000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#openrouter/auto", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 16384, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#openrouter/auto", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "auto_routing", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://openrouter.ai/api/v1/models", + "verification": "tool-use-officially-verified-other-tags-inherited" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#openrouter/auto", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/openrouter.json#openrouter/auto", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "OpenRouter Models API 的 prompt/completion=-1 为动态路由占位值，实际价格取决于选中的模型，不代表免费。", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "OpenRouter Auto", + "description": "OpenRouter 自动路由，智能选择最优模型" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/openrouter.json#openrouter/auto", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/providers/openrouter.json" + } + }, + { + "id": "provider-perplexity-001", + "category": "providers", + "connection": { + "id": "provider-perplexity-001", + "provider": "perplexity", + "label": "Perplexity Sonar", + "baseUrl": "https://api.perplexity.ai", + "apiFormat": "openai-completions", + "apiKeyRef": "perplexity", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat" + ] + }, + "models": [ + { + "apiModelId": "sonar-pro", + "modelRef": "perplexity:sonar-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 15, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.requestPricingPer1k", + "amount": { + "low": 6, + "medium": 10, + "high": 14 + }, + "currency": "USD", + "unit": "thousand-requests", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Total Sonar API cost includes token costs plus a request fee based on search context size.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Sonar Pro", + "description": "Perplexity 旗舰搜索增强模型，内置实时联网搜索和引用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/perplexity.json#sonar-pro" + } + }, + { + "apiModelId": "sonar-reasoning-pro", + "modelRef": "perplexity:sonar-reasoning-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.requestPricingPer1k", + "amount": { + "low": 6, + "medium": 10, + "high": 14 + }, + "currency": "USD", + "unit": "thousand-requests", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Total Sonar API cost includes token costs plus a request fee based on search context size.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Sonar Reasoning Pro", + "description": "Perplexity 深度推理模型，内置联网搜索和深度思考" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/perplexity.json#sonar-reasoning-pro" + } + }, + { + "apiModelId": "sonar", + "modelRef": "perplexity:sonar", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.requestPricingPer1k", + "amount": { + "low": 5, + "medium": 8, + "high": 12 + }, + "currency": "USD", + "unit": "thousand-requests", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Total Sonar API cost includes token costs plus a request fee based on search context size.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Sonar", + "description": "Perplexity 轻量搜索模型，低成本联网搜索" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/perplexity.json#sonar" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/perplexity.json" + } + }, + { + "id": "provider-siliconflow-001", + "category": "providers", + "connection": { + "id": "provider-siliconflow-001", + "provider": "siliconflow", + "label": "硅基流动", + "baseUrl": "https://api.siliconflow.cn/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "siliconflow", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "embedding" + ] + }, + "models": [ + { + "apiModelId": "Qwen/Qwen3-Coder-480B-A35B-Instruct", + "modelRef": "qwen:qwen3-coder-480b", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/siliconflow.json#Qwen/Qwen3-Coder-480B-A35B-Instruct", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3 Coder 480B (硅基)", + "description": "Qwen3-Coder-480B-A35B-Instruct，480B MoE（35B 激活），原生 256K 上下文，可通过外推扩展至 1M，面向 Agentic Coding" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/siliconflow.json#Qwen/Qwen3-Coder-480B-A35B-Instruct" + } + }, + { + "apiModelId": "Qwen/Qwen3-235B-A22B-Instruct-2507", + "modelRef": "qwen:qwen3-235b", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/siliconflow.json#Qwen/Qwen3-235B-A22B-Instruct-2507", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 10, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3 235B (硅基)", + "description": "Qwen3-235B-A22B-Instruct-2507，235B MoE（22B 激活）通用指令模型，262K 上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/siliconflow.json#Qwen/Qwen3-235B-A22B-Instruct-2507" + } + }, + { + "apiModelId": "BAAI/bge-m3", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "聚合平台请求ID有独立托管合同；没有该精确ID已证alias时保留接入本地数值/能力，不通过剥前缀绑定原厂。" + }, + "overrides": { + "serviceType": { + "value": [ + "embedding" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/siliconflow.json#BAAI/bge-m3", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/providers/siliconflow.json#BAAI/bge-m3", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 0, + "reason": "preserved-access-contract", + "evidence": "compute/providers/siliconflow.json#BAAI/bge-m3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "text_embedding", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/siliconflow.json#BAAI/bge-m3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 0, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "BGE-M3 (硅基)", + "description": "多语言 Embedding 模型，支持 8K 上下文，免费版" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/siliconflow.json#BAAI/bge-m3", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/providers/siliconflow.json" + } + }, + { + "id": "provider-stability-001", + "category": "providers", + "connection": { + "id": "provider-stability-001", + "provider": "stability", + "label": "Stability SDXL", + "baseUrl": "https://api.stability.ai/v2beta", + "apiFormat": "stability-rest-api", + "apiKeyRef": "stability", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "image_gen" + ] + }, + "models": [ + { + "apiModelId": "stable-diffusion-3.5-large", + "modelRef": "stability:stable-diffusion-3.5-large", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Stable Diffusion 3.5 Large", + "description": "Stability AI 旗舰文生图模型" + }, + "recordedExtra": { + "supportedImageSizes": [ + "1024x1024", + "1152x896", + "896x1152", + "1216x832", + "832x1216" + ], + "supportedStyles": [ + "photographic", + "digital-art", + "anime", + "comic-book" + ] + }, + "sourceRecord": "compute/providers/stability.json#stable-diffusion-3.5-large", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "stability-rest-api", + "adapterId": "required-adapter:image_gen:stability" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/stability.json" + } + }, + { + "id": "provider-tencent-001", + "category": "providers", + "connection": { + "id": "provider-tencent-001", + "provider": "tencent", + "label": "腾讯混元", + "baseUrl": "https://api.hunyuan.cloud.tencent.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "tencent", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ] + }, + "models": [ + { + "apiModelId": "hunyuan-2.0-thinking-20251109", + "modelRef": "tencent:hunyuan-2.0-thinking-20251109", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-2.0-thinking-20251109", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-2.0-thinking-20251109", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3.975, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 15.9, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32768, + "inputPrice": 3.975, + "outputPrice": 15.9 + }, + { + "maxInputTokens": 131072, + "inputPrice": 5.3, + "outputPrice": 21.2 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 2.0 Think", + "description": "腾讯混元最新推理模型，128K输入/64K输出，MoE架构406B总参数" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/tencent.json#hunyuan-2.0-thinking-20251109" + } + }, + { + "apiModelId": "hunyuan-2.0-instruct-20251111", + "modelRef": "tencent:hunyuan-2.0-instruct-20251111", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-2.0-instruct-20251111", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3.18, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 7.95, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32768, + "inputPrice": 3.18, + "outputPrice": 7.95 + }, + { + "maxInputTokens": 131072, + "inputPrice": 4.505, + "outputPrice": 11.13 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 2.0 Instruct", + "description": "腾讯混元 2.0 指令模型，最大输入128K/最大输出16K，支持联网搜索和 Function Calling" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/tencent.json#hunyuan-2.0-instruct-20251111" + } + }, + { + "apiModelId": "hunyuan-turbos-latest", + "modelRef": "tencent:hunyuan-turbos-latest", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-turbos-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 Turbo S", + "description": "腾讯混元高速模型，32K输入/16K输出，支持推理和工具调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/tencent.json#hunyuan-turbos-latest" + } + }, + { + "apiModelId": "hunyuan-t1-latest", + "modelRef": "tencent:hunyuan-t1-latest", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-t1-latest", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-t1-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 T1", + "description": "腾讯混元 T1 深度推理模型，256K 上下文，专攻数学/代码/复杂推理" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/tencent.json#hunyuan-t1-latest" + } + }, + { + "apiModelId": "hunyuan-t1-vision", + "modelRef": "tencent:hunyuan-t1-vision", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-t1-vision", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-t1-vision", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 20, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 T1 Vision", + "description": "腾讯混元 T1 视觉推理模型，支持图片理解 + 深度推理" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/tencent.json#hunyuan-t1-vision" + } + }, + { + "apiModelId": "hunyuan-turbos-vision", + "modelRef": "tencent:hunyuan-turbos-vision", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/tencent.json#hunyuan-turbos-vision", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 Turbo S Vision", + "description": "腾讯混元 Turbo S 视觉模型，支持图片理解，高速推理" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/tencent.json#hunyuan-turbos-vision" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/tencent.json" + } + }, + { + "id": "provider-volcengine-001", + "category": "providers", + "connection": { + "id": "provider-volcengine-001", + "provider": "volcengine", + "label": "火山 豆包", + "baseUrl": "https://ark.cn-beijing.volces.com/api/v3", + "apiFormat": "openai-completions", + "apiKeyRef": "volcengine", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning", + "vision", + "embedding", + "voice_clone", + "realtime_voice", + "simultaneous_interpret", + "translation", + "image_gen", + "video_gen" + ] + }, + "models": [ + { + "apiModelId": "doubao-seed-evolving", + "modelRef": "volcengine:doubao-seed-evolving", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Doubao Seed Evolving", + "description": "Seed 系列持续迭代模型，通过固定 ID 获取最新版本，2026-09-09 更新" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-evolving" + } + }, + { + "apiModelId": "doubao-seed-2.1-pro", + "modelRef": "volcengine:doubao-seed-2.1-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-pro", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-pro", + "verification": "inherited-not-reverified" + } + }, + "request": { + "upstreamModelId": "doubao-seed-2-1-pro-260628" + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 30, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1.2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePricePerMillionTokenHour", + "amount": 0.017, + "currency": "CNY", + "unit": "million-token-hour", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.1 Pro", + "description": "火山方舟豆包 Seed 2.1 Pro，生产级 Coding、Agent 与多模态旗舰模型" + }, + "recordedExtra": { + "supportsResponsesApi": true, + "supportsChatApi": true + }, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-2.1-pro" + } + }, + { + "apiModelId": "doubao-seed-2.1-turbo", + "modelRef": "volcengine:doubao-seed-2.1-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "multilingual", + "long_context", + "tool_use", + "agent", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": { + "upstreamModelId": "doubao-seed-2-1-turbo-260628" + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 15, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.6, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePricePerMillionTokenHour", + "amount": 0.017, + "currency": "CNY", + "unit": "million-token-hour", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.1 Turbo", + "description": "火山方舟豆包 Seed 2.1 Turbo，高吞吐低成本生产模型，适合规模化 Agent 和 Coding 调用" + }, + "recordedExtra": { + "supportsResponsesApi": true, + "supportsChatApi": true + }, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-2.1-turbo" + } + }, + { + "apiModelId": "doubao-seed-2.0-pro", + "modelRef": "volcengine:doubao-seed-2.0-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.0-pro", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 3.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 128000, + "inputPrice": 4.8, + "outputPrice": 24 + }, + { + "maxInputTokens": 256000, + "inputPrice": 9.6, + "outputPrice": 48 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 2.0 Pro", + "description": "火山引擎豆包 Seed 2.0 旗舰模型，256K上下文，支持复杂推理、Agent 和多模态理解" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-2.0-pro" + } + }, + { + "apiModelId": "doubao-seed-2.0-lite", + "modelRef": "volcengine:doubao-seed-2.0-lite", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "maxOutputTokens": { + "value": 128000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.0-lite", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.0-lite", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3.6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.6, + "outputPrice": 3.6 + }, + { + "maxInputTokens": 128000, + "inputPrice": 0.9, + "outputPrice": 5.4 + }, + { + "maxInputTokens": 256000, + "inputPrice": 1.8, + "outputPrice": 10.8 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 2.0 Lite", + "description": "火山引擎豆包 Seed 2.0 全模态均衡模型，256K上下文，支持视频/图像/音频/文本统一理解" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-2.0-lite" + } + }, + { + "apiModelId": "doubao-seed-2.0-mini", + "modelRef": "volcengine:doubao-seed-2.0-mini", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "maxOutputTokens": { + "value": 128000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.0-mini", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.0-mini", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.2, + "outputPrice": 2 + }, + { + "maxInputTokens": 128000, + "inputPrice": 0.4, + "outputPrice": 4 + }, + { + "maxInputTokens": 256000, + "inputPrice": 0.8, + "outputPrice": 8 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 2.0 Mini", + "description": "火山引擎豆包 Seed 2.0 全模态轻量模型，256K上下文，低延迟高并发，支持全模态理解" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-2.0-mini" + } + }, + { + "apiModelId": "doubao-seed-2.0-code", + "modelRef": "volcengine:doubao-seed-2.0-code", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-2.0-code", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 3.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 128000, + "inputPrice": 4.8, + "outputPrice": 24 + }, + { + "maxInputTokens": 256000, + "inputPrice": 9.6, + "outputPrice": 48 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 2.0 Code", + "description": "火山引擎豆包 Seed 2.0 编程模型，面向 Agentic Coding 场景优化" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-2.0-code" + } + }, + { + "apiModelId": "doubao-seed-1.8", + "modelRef": "volcengine:doubao-seed-1.8", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.8", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.8, + "outputPrice": 8, + "pricingNotes": "Short outputs up to 0.2K tokens are priced at 2 CNY/M output tokens." + }, + { + "maxInputTokens": 128000, + "inputPrice": 1.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 256000, + "inputPrice": 2.4, + "outputPrice": 24 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-1.8", + "description": "火山引擎豆包1.8旗舰模型，256K上下文，支持深度思考和多模态" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-1.8" + } + }, + { + "apiModelId": "doubao-seed-1.6", + "modelRef": "volcengine:doubao-seed-1.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.6", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.8, + "outputPrice": 8, + "pricingNotes": "Short outputs up to 0.2K tokens are priced at 2 CNY/M output tokens." + }, + { + "maxInputTokens": 128000, + "inputPrice": 1.2, + "outputPrice": 16 + }, + { + "maxInputTokens": 256000, + "inputPrice": 2.4, + "outputPrice": 24 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-1.6", + "description": "火山引擎豆包1.6旗舰模型，256K上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-1.6" + } + }, + { + "apiModelId": "doubao-seed-1.6-thinking", + "modelRef": "volcengine:doubao-seed-1.6-thinking", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.6-thinking", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.6-thinking", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-1.6 思考版", + "description": "火山引擎豆包1.6深度思考模型，256K上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-1.6-thinking" + } + }, + { + "apiModelId": "doubao-seed-1.6-flash", + "modelRef": "volcengine:doubao-seed-1.6-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.6-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.15, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 1.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-1.6 Flash", + "description": "火山引擎豆包1.6快速版，成本较前代下降63%" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-1.6-flash" + } + }, + { + "apiModelId": "doubao-seed-1.6-lite", + "modelRef": "volcengine:doubao-seed-1.6-lite", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.6-lite", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "maxInputTokens": 32000, + "inputPrice": 0.3, + "outputPrice": 2.4 + }, + { + "maxInputTokens": 128000, + "inputPrice": 0.6, + "outputPrice": 4 + }, + { + "maxInputTokens": 256000, + "inputPrice": 1.2, + "outputPrice": 12 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-1.6 Lite", + "description": "火山引擎豆包1.6轻量版，极致性价比" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-1.6-lite" + } + }, + { + "apiModelId": "doubao-seed-1.6-vision", + "modelRef": "volcengine:doubao-seed-1.6-vision", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-1.6-vision", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "deprecated": true, + "deprecationNotice": "火山方舟 2026-07-06 模型列表中标记为即将下线，建议迁移到 Seed 2.1 或 Seed 2.0 系列。", + "replacedBy": "doubao-seed-2.1-turbo" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-1.6 视觉版", + "description": "火山引擎豆包1.6视觉模型，256K上下文，支持视频理解" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-1.6-vision" + } + }, + { + "apiModelId": "doubao-seed-code", + "modelRef": "volcengine:doubao-seed-code", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.7, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seed-code", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed-Code", + "description": "火山引擎豆包编程专用模型，256K上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-seed-code" + } + }, + { + "apiModelId": "deepseek-v3.2", + "modelRef": "volcengine:deepseek-v3.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#deepseek-v3.2", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V3.2 (火山引擎)", + "description": "火山引擎托管的DeepSeek V3.2，128K上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#deepseek-v3.2" + } + }, + { + "apiModelId": "deepseek-r1", + "modelRef": "volcengine:deepseek-r1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 16, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek R1 (火山引擎)", + "description": "火山引擎托管的DeepSeek R1推理模型，128K上下文" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#deepseek-r1" + } + }, + { + "apiModelId": "doubao-embedding", + "modelRef": "volcengine:doubao-embedding", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Embedding", + "description": "火山引擎豆包标准向量模型，默认2048维度，可选512/1024/2048维度" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-embedding" + } + }, + { + "apiModelId": "doubao-embedding-large", + "modelRef": "volcengine:doubao-embedding-large", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Embedding Large", + "description": "火山引擎豆包大号向量模型，默认2048维度，可选512/1024/2048/4096维度" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#doubao-embedding-large" + } + }, + { + "apiModelId": "volc-mega-tts-clone", + "modelRef": "volcengine:volc-mega-tts-clone", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "火山声音复刻", + "description": "火山引擎声音复刻，支持极少样本" + }, + "recordedExtra": { + "maxCloneSamples": 3, + "minCloneDuration": 5 + }, + "sourceRecord": "compute/providers/volcengine.json#volc-mega-tts-clone" + } + }, + { + "apiModelId": "volc-realtime-voice", + "modelRef": "volcengine:volc-realtime-voice", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "火山实时语音", + "description": "火山引擎实时语音交互" + }, + "recordedExtra": { + "supportedModes": [ + "speech-to-speech" + ], + "latencyMs": 200 + }, + "sourceRecord": "compute/providers/volcengine.json#volc-realtime-voice" + } + }, + { + "apiModelId": "volc-simultaneous", + "modelRef": "volcengine:volc-simultaneous", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "火山同声传译", + "description": "火山引擎同声传译，支持中英双向" + }, + "recordedExtra": { + "streamingSupported": true + }, + "sourceRecord": "compute/providers/volcengine.json#volc-simultaneous" + } + }, + { + "apiModelId": "volc-translation", + "modelRef": "volcengine:volc-translation", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "火山翻译", + "description": "火山引擎机器翻译，支持多语种互译" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/volcengine.json#volc-translation" + } + }, + { + "apiModelId": "doubao-seedream-4-0-250828", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "image_gen" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seedream-4-0-250828", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "image_generation", + "image_to_image", + "multi_image_fusion", + "chinese_optimized", + "high_quality" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seedream-4-0-250828", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seedream 4.0（文生图/图生图）", + "description": "字节 Seedream 图像模型，支持文生图、图生图（img2img）、局部重绘与多图融合（最多 10 张参考图）；OpenAI 风格 /images/generations 同步返回" + }, + "recordedExtra": { + "endpoint": "/images/generations", + "supportedSizes": [ + "1K", + "2K", + "4K" + ], + "maxFusionImages": 10 + }, + "sourceRecord": "compute/providers/volcengine.json#doubao-seedream-4-0-250828", + "automaticRouting": false + } + }, + { + "apiModelId": "doubao-seedance-2-0-260128", + "modelRef": "volcengine:doubao-seedance-2.0", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "text_to_video", + "image_to_video", + "first_last_frame", + "reference_video", + "reference_audio", + "chinese_optimized", + "high_quality" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/volcengine.json#doubao-seedance-2-0-260128", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seedance 2.0（文生/图生视频）", + "description": "字节 Seedance 视频模型，支持文生视频、图生视频、多图参考、首尾帧、参考视频/音频（多模态），异步任务" + }, + "recordedExtra": { + "endpoint": "/contents/generations/tasks", + "maxVideoDuration": 15, + "supportedResolutions": [ + "720p", + "1080p" + ], + "maxReferenceImages": 9, + "asyncMode": true + }, + "sourceRecord": "compute/providers/volcengine.json#doubao-seedance-2-0-260128" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/volcengine.json" + } + }, + { + "id": "provider-xai-001", + "category": "providers", + "connection": { + "id": "provider-xai-001", + "provider": "xai", + "label": "xAI Grok", + "baseUrl": "https://api.x.ai/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "xai", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "USD", + "services": [ + "chat", + "reasoning" + ], + "tombstones": [ + "grok-4-1-fast-reasoning" + ] + }, + "models": [ + { + "apiModelId": "grok-4.7", + "modelRef": "xai:grok-4.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "structured_output", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.7", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh" + ], + "defaultEffort": "high" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 6, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.5, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "Prices are per 1M tokens below 200K prompt tokens; above 200K, input and cache rates are 2x and output rates are 2x. No text output limit is documented.", + "currency": "USD", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Grok 4.7", + "description": "xAI 面向编程与智能体任务的旗舰模型，500K 上下文，支持图像输入和可配置推理" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xai.json#grok-4.7" + } + }, + { + "apiModelId": "grok-4.3", + "modelRef": "xai:grok-4-3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "vision", + "tool_use", + "structured_output", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.3", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.3", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.3", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "none", + "low", + "medium", + "high" + ], + "defaultEffort": "low" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Grok 4.3", + "description": "xAI Grok 4.3 通用模型，1M 上下文，支持可配置推理、函数调用和结构化输出" + }, + "recordedExtra": { + "aliases": [ + "grok-4.3-latest", + "grok-latest" + ] + }, + "sourceRecord": "compute/providers/xai.json#grok-4.3" + } + }, + { + "apiModelId": "grok-build-0.1", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-build-0.1", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 256000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-build-0.1", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "code", + "agent", + "tool_use", + "structured_output" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-build-0.1", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-build-0.1", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-build-0.1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Grok Build 0.1", + "description": "xAI 面向 agentic coding 工作流的专用代码模型" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xai.json#grok-build-0.1", + "automaticRouting": false + } + }, + { + "apiModelId": "grok-4.20-0309-reasoning", + "modelRef": "xai:grok-4.20-0309-reasoning", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 1000000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.20-0309-reasoning", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "vision", + "tool_use", + "structured_output", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.20-0309-reasoning", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xai.json#grok-4.20-0309-reasoning", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1.25, + "currency": "USD", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.5, + "currency": "USD", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.2, + "currency": "USD", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown", + "legacy": true + }, + "metadata": { + "attributes": { + "displayName": "Grok 4.20 Reasoning (2026-03-09)", + "description": "xAI Grok 4.20 固定版本推理模型，当前价格与 Grok 4.3 一致" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xai.json#grok-4.20-0309-reasoning" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/xai.json" + } + }, + { + "id": "provider-xiaomi-001", + "category": "providers", + "connection": { + "id": "provider-xiaomi-001", + "provider": "xiaomi", + "label": "小米 MiMo", + "baseUrl": "https://api.xiaomimimo.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "xiaomi", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "brandGroup": "xiaomi", + "priceCurrency": "CNY", + "accessMode": "api", + "services": [ + "chat", + "reasoning", + "vision", + "tts", + "asr" + ] + }, + "models": [ + { + "apiModelId": "mimo-v2.6-pro", + "modelRef": "xiaomi:mimo-v2.6-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.025, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "CNY per 1M tokens on the Xiaomi MiMo API.", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.6-Pro", + "description": "小米全模态旗舰推理模型，适合复杂项目与长程智能体任务" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.6-pro" + } + }, + { + "apiModelId": "mimo-v2.6-flash", + "modelRef": "xiaomi:mimo-v2.6-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.02, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "CNY per 1M tokens on the Xiaomi MiMo API.", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.6-Flash", + "description": "小米全模态高效推理模型，适合高频调用与规模化任务" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.6-flash" + } + }, + { + "apiModelId": "mimo-v2.6-pro-ultraspeed", + "modelRef": "xiaomi:mimo-v2.6-pro-ultraspeed", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 30, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 60, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.25, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "CNY per 1M tokens on the Xiaomi MiMo API; UltraSpeed requires custom rate limits.", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.6-Pro-UltraSpeed", + "description": "MiMo-V2.6-Pro 的超高速服务，需向小米申请定制配额" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.6-pro-ultraspeed" + } + }, + { + "apiModelId": "mimo-x-flash-preview", + "modelRef": "xiaomi:mimo-x-flash-preview", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-X-Flash Preview", + "description": "小米 MiMo-X Flash 内测模型，1M 上下文，支持图片识别" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xiaomi.json#mimo-x-flash-preview" + } + }, + { + "apiModelId": "mimo-x-pro-preview", + "modelRef": "xiaomi:mimo-x-pro-preview", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-X-Pro Preview", + "description": "小米 MiMo-X Pro 内测模型，1M 上下文，支持图片识别" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xiaomi.json#mimo-x-pro-preview" + } + }, + { + "apiModelId": "mimo-v2.5-pro", + "modelRef": "xiaomi:mimo-v2.5-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cachedInputPrice", + "amount": 0.025, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.5-Pro", + "description": "小米 MiMo-V2.5-Pro，1M 上下文，1T MoE 参数，计划于 2026-10-21 下线" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.5-pro" + } + }, + { + "apiModelId": "mimo-v2.5-tts", + "modelRef": "xiaomi:mimo-v2.5-tts", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "tts", + "style_control", + "chinese_optimized" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xiaomi.json#mimo-v2.5-tts", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.5-TTS", + "description": "小米最新高品质语音合成模型，内置多种精调音色，支持情感/语速/语调精细控制、方言和歌唱" + }, + "recordedExtra": { + "voices": [ + { + "id": "mimo_default", + "label": "MiMo 默认" + }, + { + "id": "冰糖", + "label": "冰糖" + }, + { + "id": "茉莉", + "label": "茉莉" + }, + { + "id": "苏打", + "label": "苏打" + }, + { + "id": "白桦", + "label": "白桦" + }, + { + "id": "Mia", + "label": "Mia" + }, + { + "id": "Chloe", + "label": "Chloe" + }, + { + "id": "Milo", + "label": "Milo" + }, + { + "id": "Dean", + "label": "Dean" + } + ], + "audioFormats": [ + "wav", + "mp3", + "pcm16le" + ], + "sampleRate": 24000, + "endpoint": "/chat/completions", + "authHeader": "api-key" + }, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.5-tts" + } + }, + { + "apiModelId": "mimo-v2.5-asr", + "modelRef": "xiaomi:mimo-v2.5-asr", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.5-ASR", + "description": "小米 MiMo 语音识别模型，支持中英文提示与整段 WAV/MP3 识别" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.5-asr" + } + }, + { + "apiModelId": "mimo-v2.5-tts-voicedesign", + "modelRef": "xiaomi:mimo-v2.5-tts-voicedesign", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "tts", + "voice_design" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xiaomi.json#mimo-v2.5-tts-voicedesign", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.5-TTS-VoiceDesign", + "description": "语音设计模型，通过一句话描述即可生成全新音色" + }, + "recordedExtra": { + "audioFormats": [ + "mp3", + "wav" + ], + "sampleRate": 24000, + "endpoint": "/chat/completions", + "authHeader": "api-key" + }, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.5-tts-voicedesign", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:tts:xiaomi" + } + }, + { + "apiModelId": "mimo-v2.5-tts-voiceclone", + "modelRef": "xiaomi:mimo-v2.5-tts-voiceclone", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "tts" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xiaomi.json#mimo-v2.5-tts-voiceclone", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "tts", + "voice_clone" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/xiaomi.json#mimo-v2.5-tts-voiceclone", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiMo-V2.5-TTS-VoiceClone", + "description": "语音克隆模型，少量样本即可高保真复刻目标音色" + }, + "recordedExtra": { + "audioFormats": [ + "mp3", + "wav" + ], + "sampleRate": 24000, + "endpoint": "/chat/completions", + "authHeader": "api-key" + }, + "sourceRecord": "compute/providers/xiaomi.json#mimo-v2.5-tts-voiceclone", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:tts:xiaomi" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/xiaomi.json" + } + }, + { + "id": "provider-xunfei-x2-001", + "category": "providers", + "connection": { + "id": "provider-xunfei-x2-001", + "provider": "xunfei", + "label": "讯飞星火 X2", + "baseUrl": "https://spark-api-open.xf-yun.com/x2", + "apiFormat": "openai-completions", + "apiKeyRef": "xunfei", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ] + }, + "models": [ + { + "apiModelId": "spark-x", + "modelRef": "xunfei:spark-x", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xunfei-x2.json#spark-x", + "verification": "inherited-not-reverified" + }, + "maxInputTokens": { + "value": 65536, + "reason": "V2将旧contextWindow65536拆为独立输入上限，总窗口unknown；V1冻结投影保留旧字段待客户端正式发布。", + "evidence": "https://www.xfyun.cn/doc/spark/X1http.html", + "verification": "officially-verified-independent-budget" + }, + "contextWindow": { + "value": null, + "reason": "V2将旧contextWindow65536拆为独立输入上限，总窗口unknown；V1冻结投影保留旧字段待客户端正式发布。", + "evidence": "https://www.xfyun.cn/doc/spark/X1http.html", + "verification": "officially-verified-independent-budget" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "V2将旧contextWindow65536拆为独立输入上限，总窗口unknown；V1冻结投影保留旧字段待客户端正式发布。", + "evidence": "https://www.xfyun.cn/doc/spark/X1http.html", + "verification": "officially-verified-independent-budget" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricingNote", + "amount": "官方公开文档说明模型价格以控制台及实际购买页为准，未公开固定 token 单价", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "讯飞星火 X2", + "description": "讯飞深度推理模型（X2），输入64K/输出128K，支持思考模式与函数调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xunfei-x2.json#spark-x", + "compatibilityValues": { + "contextWindow": 65536 + }, + "independentBudgetEvidence": { + "file": "compute/providers/xunfei-x2.json", + "apiModelId": "spark-x", + "maxInputTokens": 65536, + "contextWindow": null, + "maxOutputTokens": 131072, + "evidence": "https://www.xfyun.cn/doc/spark/X1http.html", + "checkedAt": "2026-10-03", + "contentSha256": "324d06b132ff4ab16a0543157467283dc789153303cf456b44af45d1fc6f1548", + "unitConversion": "官网明文输入64K/输出128K；同页max_tokens最大/默认131072，128K按1024解释；64K转换为65536并明确为换算，不宣称官网明文整数。", + "semanticDecision": "V2将旧contextWindow65536拆为独立输入上限，总窗口unknown；V1冻结投影保留旧字段待客户端正式发布。" + } + } + } + ], + "metadata": { + "sourceFile": "compute/providers/xunfei-x2.json" + } + }, + { + "id": "provider-xunfei-001", + "category": "providers", + "connection": { + "id": "provider-xunfei-001", + "provider": "xunfei", + "label": "讯飞星火", + "baseUrl": "https://spark-api-open.xf-yun.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "xunfei", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "reasoning" + ], + "tombstones": [ + "spark-x" + ] + }, + "models": [ + { + "apiModelId": "4.0Ultra", + "modelRef": "xunfei:4.0Ultra", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "defaultTopP": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/xunfei.json#4.0Ultra", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "extra.pricingNote", + "amount": "官方公开文档说明模型价格以控制台及实际购买页为准，未公开固定 token 单价", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "讯飞星火 4.0 Ultra", + "description": "讯飞星火最强通用模型，支持推理、代码和工具调用" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/xunfei.json#4.0Ultra" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/xunfei.json" + } + }, + { + "id": "provider-zhipu-embedding-001", + "category": "providers", + "connection": { + "id": "provider-zhipu-embedding-001", + "provider": "zhipu-embedding", + "brandGroup": "zhipu", + "label": "智谱 Embedding", + "baseUrl": "https://open.bigmodel.cn/api/paas/v4", + "apiFormat": "openai-completions", + "apiKeyRef": "zhipu", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "embedding" + ] + }, + "models": [ + { + "apiModelId": "embedding-3", + "modelRef": "zhipu:embedding-3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "智谱 embedding-3", + "description": "智谱嵌入模型v3，支持自定义维度；模型上下文窗口 8K，单条输入最多 3072 tokens。走 OpenAI 兼容 /embeddings 端点（Anthropic 协议不提供 embeddings）" + }, + "recordedExtra": { + "maxInputTokensPerItem": 3072, + "maxBatchItems": 64, + "vectorDimensions": "256-2048" + }, + "sourceRecord": "compute/providers/zhipu-embedding.json#embedding-3" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/zhipu-embedding.json" + } + }, + { + "id": "provider-zhipu-001", + "category": "providers", + "connection": { + "id": "provider-zhipu-001", + "provider": "zhipu", + "label": "智谱 GLM", + "baseUrl": "https://open.bigmodel.cn/api/anthropic", + "apiFormat": "anthropic-messages", + "apiKeyRef": "zhipu", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "priceCurrency": "CNY", + "services": [ + "chat", + "vision", + "reasoning", + "fast" + ], + "tombstones": [ + "embedding-3", + "glm-4.7-thinking", + "glm-z1-air", + "glm-z1-airx", + "glm-z1-flash", + "glm-z1-flashx", + "glm-4-0520" + ] + }, + "models": [ + { + "apiModelId": "glm-5.3-flashx", + "modelRef": "zhipu:glm-5.3-flashx", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "adaptiveThinking": true, + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 7, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.57, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.3-FlashX", + "description": "Z.ai GLM-5.3-FlashX，高速多模态推理模型，100 万上下文，始终开启推理" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/zhipu.json#glm-5.3-flashx" + } + }, + { + "apiModelId": "glm-5.3-flash", + "modelRef": "zhipu:glm-5.3-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "adaptiveThinking": true, + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 2.8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.23, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.3-Flash", + "description": "智谱 GLM-5.3-Flash 原生多模态模型，混合稀疏与线性注意力，面向高效编码与长程 Agent 任务，100 万上下文；强制思考，无法关闭" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/providers/zhipu.json#glm-5.3-flash" + } + }, + { + "apiModelId": "glm-5.3", + "modelRef": "zhipu:glm-5.3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "adaptiveThinking": true, + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 28, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingNotes", + "amount": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。", + "currency": "CNY", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.3", + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/providers/zhipu.json#glm-5.3" + } + }, + { + "apiModelId": "glm-5.2", + "modelRef": "zhipu:glm-5.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1000000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "math", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.2", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + }, + "reasoning": { + "supportedEfforts": [ + "max", + "xhigh", + "high", + "medium", + "low", + "minimal", + "none" + ], + "defaultEffort": "max" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 28, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.2", + "description": "智谱 GLM-5.2 最新旗舰模型，面向复杂长程任务与项目级工程交付，1M 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/zhipu.json#glm-5.2" + } + }, + { + "apiModelId": "glm-5.1", + "modelRef": "zhipu:glm-5.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.1", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.1", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5.1", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 6, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 24, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1.3, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 6, + "outputPrice": 24, + "cacheHitPrice": 1.3 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 8, + "outputPrice": 28, + "cacheHitPrice": 2 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.1", + "description": "智谱 GLM-5.1 旗舰模型，面向长程 Agentic Coding 与 Autonomous Agent 场景，200K 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/zhipu.json#glm-5.1" + } + }, + { + "apiModelId": "glm-5-turbo", + "modelRef": "zhipu:glm-5-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5-turbo", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5-turbo", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5-turbo", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 22, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1.2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 5, + "outputPrice": 22, + "cacheHitPrice": 1.2 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 7, + "outputPrice": 26, + "cacheHitPrice": 1.8 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5-Turbo", + "description": "智谱 GLM-5-Turbo，高阶快速模型，面向长链路任务优化，200K 上下文，128K 最大输出" + }, + "recordedExtra": { + "decodeSpeed": "40-60 tokens/s" + }, + "sourceRecord": "compute/providers/zhipu.json#glm-5-turbo" + } + }, + { + "apiModelId": "glm-5", + "modelRef": "zhipu:glm-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 4, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 18, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 4, + "outputPrice": 18, + "cacheHitPrice": 1 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 6, + "outputPrice": 22, + "cacheHitPrice": 1.5 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5", + "description": "智谱 GLM-5，面向 Agentic Engineering 的旗舰基座模型，200K 上下文，128K 最大输出" + }, + "recordedExtra": { + "decodeSpeed": "30-50 tokens/s" + }, + "sourceRecord": "compute/providers/zhipu.json#glm-5" + } + }, + { + "apiModelId": "glm-4.7", + "modelRef": "zhipu:glm-4.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled", + "mode": "forced" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.4, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "input_length_k_tokens: [0, 32); output_length_k_tokens: [0, 0.2)", + "inputPrice": 2, + "outputPrice": 8, + "cacheHitPrice": 0.4 + }, + { + "condition": "input_length_k_tokens: [0, 32); output_length_k_tokens: [0.2+)", + "inputPrice": 3, + "outputPrice": 14, + "cacheHitPrice": 0.6 + }, + { + "condition": "input_length_k_tokens: [32, 200)", + "inputPrice": 4, + "outputPrice": 16, + "cacheHitPrice": 0.8 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.7", + "description": "智谱 GLM-4.7，高智能模型，通用对话、推理与智能体能力全面升级，200K 上下文，128K 最大输出" + }, + "recordedExtra": { + "decodeSpeed": "30-50 tokens/s" + }, + "sourceRecord": "compute/providers/zhipu.json#glm-4.7" + } + }, + { + "apiModelId": "glm-4.7-flashx", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7-flashx", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 200000, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7-flashx", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7-flashx", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "fast", + "long_context", + "multilingual", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://docs.bigmodel.cn/cn/guide/models/text/glm-4.7.md", + "verification": "tool-use-officially-verified-other-tags-inherited" + }, + "defaultTemperature": { + "value": 1, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7-flashx", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.7-flashx", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.1, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.7-FlashX", + "description": "智谱 GLM-4.7-FlashX 轻量高速模型，适用于中文写作、翻译、角色扮演等通用场景，200K 上下文，128K 最大输出" + }, + "recordedExtra": { + "decodeSpeed": "30-40 tokens/s" + }, + "sourceRecord": "compute/providers/zhipu.json#glm-4.7-flashx", + "automaticRouting": false + } + }, + { + "apiModelId": "glm-5v-turbo", + "modelRef": "zhipu:glm-5v-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5v-turbo", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-5v-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 5, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 22, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 1.2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 5, + "outputPrice": 22, + "cacheHitPrice": 1.2 + }, + { + "condition": "input_length_k_tokens: [32+)", + "inputPrice": 7, + "outputPrice": 26, + "cacheHitPrice": 1.8 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5V-Turbo", + "description": "智谱首个多模态 Agent 基座模型，支持视频、图像、文本和文件输入，200K 上下文，128K 最大输出" + }, + "recordedExtra": { + "decodeSpeed": "30-50 tokens/s", + "inputModalities": [ + "image", + "video", + "file", + "text" + ] + }, + "sourceRecord": "compute/providers/zhipu.json#glm-5v-turbo" + } + }, + { + "apiModelId": "glm-4.6v", + "modelRef": "zhipu:glm-4.6v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "vision", + "video_understanding", + "image_understanding", + "file_understanding", + "long_context", + "tool_use", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.6v", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.6v", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 1, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 3, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.2, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheStoragePrice", + "amount": 0, + "currency": "CNY", + "unit": "source-unit-unresolved", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.pricingTiers", + "amount": [ + { + "condition": "input_length_k_tokens: [0, 32)", + "inputPrice": 1, + "outputPrice": 3, + "cacheHitPrice": 0.2 + }, + { + "condition": "input_length_k_tokens: [32, 128)", + "inputPrice": 2, + "outputPrice": 6, + "cacheHitPrice": 0.4 + } + ], + "currency": "CNY", + "unit": "tiered-source-unit", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.6V", + "description": "智谱 GLM-4.6V 多模态模型，支持图像、视频、文本和文件理解，128K 上下文，32K 最大输出" + }, + "recordedExtra": { + "decodeSpeed": "30-50 tokens/s", + "inputModalities": [ + "image", + "video", + "file", + "text" + ] + }, + "sourceRecord": "compute/providers/zhipu.json#glm-4.6v" + } + }, + { + "apiModelId": "glm-4.6", + "modelRef": "zhipu:glm-4.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.6", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.6", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "deep_thinking", + "long_context", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.6", + "verification": "inherited-not-reverified" + }, + "defaultTopP": { + "value": 0.95, + "reason": "preserved-access-contract", + "evidence": "compute/providers/zhipu.json#glm-4.6", + "verification": "inherited-not-reverified" + } + }, + "request": { + "thinking": { + "default": "enabled" + } + }, + "billing": [ + { + "sourceKey": "inputPrice", + "amount": 2, + "currency": "CNY", + "unit": "million-tokens", + "direction": "input", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "outputPrice", + "amount": 8, + "currency": "CNY", + "unit": "million-tokens", + "direction": "output", + "verification": "inherited-not-reverified" + }, + { + "sourceKey": "extra.cacheHitPrice", + "amount": 0.5, + "currency": "CNY", + "unit": "million-tokens", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.6", + "description": "智谱 GLM-4.6，擅长高级编码、复杂推理与工具调用，200K 上下文，128K 最大输出" + }, + "recordedExtra": {}, + "sourceRecord": "compute/providers/zhipu.json#glm-4.6" + } + } + ], + "metadata": { + "sourceFile": "compute/providers/zhipu.json" + } + }, + { + "id": "provider-baidu-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-baidu-coding-001", + "provider": "baidu", + "brandGroup": "baidu", + "label": "百度千帆 Coding Plan", + "baseUrl": "https://qianfan.baidubce.com/v2/coding", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://console.bce.baidu.com/qianfan/codingplan" + } + } + }, + "models": [ + { + "apiModelId": "qianfan-code-latest", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#qianfan-code-latest", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#qianfan-code-latest", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Auto (自动路由)", + "description": "千帆 Coding Plan 自动路由模型，后端自动选择最优模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#qianfan-code-latest", + "automaticRouting": false + } + }, + { + "apiModelId": "deepseek-v3.2", + "modelRef": "volcengine:deepseek-v3.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#deepseek-v3.2", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V3.2", + "description": "DeepSeek V3.2（千帆 Coding Plan 渠道），倍率 ×1", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#deepseek-v3.2" + } + }, + { + "apiModelId": "deepseek-v4-flash", + "modelRef": "deepseek:deepseek-v4-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#deepseek-v4-flash", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#deepseek-v4-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Flash", + "description": "DeepSeek V4 Flash（千帆 Coding Plan 渠道），倍率 ×1", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#deepseek-v4-flash" + } + }, + { + "apiModelId": "ernie-4.5-turbo-20260402", + "modelRef": "baidu:ernie-4.5-turbo-20260402", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "code", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#ernie-4.5-turbo-20260402", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "文心 ERNIE 4.5 Turbo", + "description": "百度 ERNIE 4.5 Turbo（千帆 Coding Plan 渠道）；套餐当前不支持图像理解", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#ernie-4.5-turbo-20260402" + } + }, + { + "apiModelId": "glm-5", + "modelRef": "zhipu:glm-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://cloud.baidu.com/doc/qianfan-docs/s/xm95lyys5", + "verification": "tool-use-officially-verified-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5", + "description": "智谱 GLM-5（千帆 Coding Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#glm-5" + } + }, + { + "apiModelId": "glm-5.1", + "modelRef": "zhipu:glm-5.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#glm-5.1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.1", + "description": "智谱 GLM-5.1（千帆 Coding Plan 渠道），高峰抵扣 ×4 / 低峰 ×3；高峰时段随流量动态调整", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#glm-5.1" + } + }, + { + "apiModelId": "kimi-k2.5", + "modelRef": "moonshot:kimi-k2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent", + "long_context", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://cloud.baidu.com/doc/qianfan-docs/s/xm95lyys5", + "verification": "tool-use-officially-verified-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.5", + "description": "Kimi K2.5（千帆 Coding Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#kimi-k2.5" + } + }, + { + "apiModelId": "kimi-k2.6", + "modelRef": "moonshot:kimi-k2.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#kimi-k2.6", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.6", + "description": "Kimi K2.6（千帆 Coding Plan 渠道），倍率 ×5", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#kimi-k2.6" + } + }, + { + "apiModelId": "minimax-m2.5", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#minimax-m2.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://cloud.baidu.com/doc/qianfan-docs/s/xm95lyys5", + "verification": "tool-use-officially-verified-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5", + "description": "MiniMax M2.5（千帆 Coding Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#minimax-m2.5", + "automaticRouting": false + } + }, + { + "apiModelId": "minimax-m2.7", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#minimax-m2.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/baidu-coding.json#minimax-m2.7", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7", + "description": "MiniMax M2.7（千帆 Coding Plan 渠道），倍率 ×2", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/baidu-coding.json#minimax-m2.7", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/baidu-coding.json" + } + }, + { + "id": "provider-dashscope-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-dashscope-coding-001", + "provider": "dashscope", + "brandGroup": "dashscope", + "label": "百炼 Coding Plan", + "baseUrl": "https://coding.dashscope.aliyuncs.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://bailian.console.aliyun.com/" + } + } + }, + "models": [ + { + "apiModelId": "qwen3.7-plus", + "modelRef": "qwen:qwen3.7-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3.7-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.7 Plus (Coding)", + "description": "百炼 Coding Plan，通义千问 3.7 Plus，多模态智能体模型，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#qwen3.7-plus" + } + }, + { + "apiModelId": "qwen3.6-plus", + "modelRef": "qwen:qwen3.6-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3.6-plus", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3.6-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.6 Plus (Coding)", + "description": "百炼 Coding Plan，通义千问 3.6 Plus，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#qwen3.6-plus" + } + }, + { + "apiModelId": "qwen3.5-plus", + "modelRef": "qwen:qwen3.5-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3.5-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.5 Plus (Coding)", + "description": "百炼 Coding Plan，通义千问 3.5 Plus，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#qwen3.5-plus" + } + }, + { + "apiModelId": "qwen3-coder-plus", + "modelRef": "qwen:qwen3-coder-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-plus", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1000000, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-plus", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 65536, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-plus", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "code", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3 Coder Plus (Coding)", + "description": "百炼 Coding Plan，Qwen3 编程增强版，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#qwen3-coder-plus" + } + }, + { + "apiModelId": "qwen3-coder-next", + "modelRef": "qwen:qwen3-coder-next", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-next", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-next", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 65536, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-next", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "code", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-coder-next", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3 Coder Next (Coding)", + "description": "百炼 Coding Plan，Qwen3 编程下一代实验模型，262K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#qwen3-coder-next" + } + }, + { + "apiModelId": "qwen3-max-2026-01-23", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-max-2026-01-23", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-max-2026-01-23", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 65536, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-max-2026-01-23", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual", + "tool_use", + "long_context", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#qwen3-max-2026-01-23", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3 Max (Coding)", + "description": "百炼 Coding Plan，Qwen3 Max 旗舰模型，262K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#qwen3-max-2026-01-23", + "automaticRouting": false + } + }, + { + "apiModelId": "kimi-k2.5", + "modelRef": "moonshot:kimi-k2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#kimi-k2.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "vision", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#kimi-k2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.5 (Coding)", + "description": "百炼 Coding Plan，月之暗面 Kimi K2.5，262K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#kimi-k2.5" + } + }, + { + "apiModelId": "glm-5", + "modelRef": "zhipu:glm-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 202752, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#glm-5", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#glm-5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#glm-5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5 (Coding)", + "description": "百炼 Coding Plan，智谱 GLM-5，200K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#glm-5" + } + }, + { + "apiModelId": "glm-4.7", + "modelRef": "zhipu:glm-4.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 202752, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://help.aliyun.com/zh/model-studio/glm-4-7", + "verification": "tool-use-officially-verified-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.7 (Coding)", + "description": "百炼 Coding Plan，智谱 GLM-4.7，200K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#glm-4.7" + } + }, + { + "apiModelId": "MiniMax-M2.5", + "modelRef": "minimax:MiniMax-M2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 196608, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-coding.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5 (Coding)", + "description": "百炼 Coding Plan，MiniMax M2.5，196K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-coding.json#MiniMax-M2.5" + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/dashscope-coding.json" + } + }, + { + "id": "provider-dashscope-token-plan-001", + "category": "coding-plans", + "connection": { + "id": "provider-dashscope-token-plan-001", + "provider": "dashscope", + "brandGroup": "dashscope", + "label": "百炼 Token Plan", + "baseUrl": "https://token-plan.cn-beijing.maas.aliyuncs.com/compatible-mode/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "token-plan", + "services": [ + "chat", + "reasoning" + ], + "tombstones": [ + "qwen3.8-max-preview", + "qwen3.5-plus", + "qwen3-coder-plus" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://bailian.console.aliyun.com/" + } + } + }, + "models": [ + { + "apiModelId": "deepseek-v4.1-flash", + "modelRef": "deepseek:deepseek-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4.1 Flash", + "description": "DeepSeek V4.1 Flash，支持文本与图像输入，100 万上下文", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#deepseek-v4.1-flash" + } + }, + { + "apiModelId": "wan2.7-image-pro", + "modelRef": "qwen:wan2.7-image-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.7 Pro", + "description": "万相最新旗舰文生图模型，支持 4K 分辨率", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#wan2.7-image-pro", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "wan2.7-image", + "modelRef": "qwen:wan2.7-image", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "通义万相 2.7", + "description": "万相最新标准文生图模型，高画质", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#wan2.7-image", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "qwen-image-3.0-pro", + "modelRef": "qwen:qwen-image-3.0-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "qwen-image-3.0-pro", + "description": "qwen-image-3.0-pro", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen-image-3.0-pro", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/media/adapters/registry.ts", + "sha256": "1140465675400f18de1eda59f12406c4c798db71da13c5db204a3f2970165829", + "assertion": "getImageAdapter/getVideoAdapter仅返回对应provider slug的映射，无通用fallback；SUPPORTED列表由该映射的Object.keys生成。" + }, + { + "path": "packages/agent-service/src/builtin-tools/generate-image.ts", + "assertion": "execute实际findProvider后以found.provider.provider取getAdapter；undefined立即return失败，不进入buildSubmit/proxy。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "no-registered-image_gen-provider-adapter", + "apiFormat": "openai-completions", + "adapterId": "required-adapter:image_gen:dashscope" + } + }, + { + "apiModelId": "happyhorse-1.1-t2v", + "modelRef": "happyhorse:happyhorse-1.1-t2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "HappyHorse 1.1 T2V", + "description": "HappyHorse 1.1 文生视频，进一步提升文本词义理解、镜头调度与动态生成表现", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#happyhorse-1.1-t2v" + } + }, + { + "apiModelId": "happyhorse-1.1-r2v", + "modelRef": "happyhorse:happyhorse-1.1-r2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "HappyHorse 1.1 R2V", + "description": "HappyHorse 1.1 参考生视频，支持最多9张图片参考，更加稳定的主体与场景参考", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#happyhorse-1.1-r2v" + } + }, + { + "apiModelId": "happyhorse-1.1-i2v", + "modelRef": "happyhorse:happyhorse-1.1-i2v", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "HappyHorse 1.1 I2V", + "description": "HappyHorse 1.1 图生视频，进一步提升画面质感、动态表现与跨片段一致性", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#happyhorse-1.1-i2v" + } + }, + { + "apiModelId": "glm-5.3", + "modelRef": "zhipu:glm-5.3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.3", + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "source": "preset" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#glm-5.3" + } + }, + { + "apiModelId": "glm-5.2", + "modelRef": "zhipu:glm-5.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "minimal", + "low", + "medium", + "high", + "xhigh" + ], + "defaultEffort": "high" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.2", + "description": "智谱 GLM-5.2 大推理模型，100 万上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#glm-5.2" + } + }, + { + "apiModelId": "deepseek-v4-pro-0813", + "modelRef": "deepseek:deepseek-v4-pro-0813", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Pro 0813", + "description": "DeepSeek V4 Pro 0813 版本，1.6T MoE 架构，100 万上下文", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#deepseek-v4-pro-0813" + } + }, + { + "apiModelId": "deepseek-v4-pro", + "modelRef": "deepseek:deepseek-v4-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Pro", + "description": "DeepSeek V4 Pro，1.6T MoE 架构，100 万上下文", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#deepseek-v4-pro" + } + }, + { + "apiModelId": "deepseek-v4-flash-0731", + "modelRef": "deepseek:deepseek-v4-flash-0731", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "reasoning": { + "supportedEfforts": [ + "high", + "max" + ], + "defaultEffort": "high" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V4 Flash 0731", + "description": "DeepSeek V4 Flash 0731 固定版本，100 万上下文", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#deepseek-v4-flash-0731" + } + }, + { + "apiModelId": "qwen3.8-max", + "modelRef": "qwen:qwen3.8-max", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.8-max", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.8-max", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.8 Max (Token Plan)", + "description": "百炼 Token Plan，通义千问 3.8 Max 正式版，支持文本、图像和视频输入", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true, + "thinkingMaxTokens": 262144, + "preserveThinkingDefault": true, + "supportsParallelToolCalls": false + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen3.8-max" + } + }, + { + "apiModelId": "qwen3.8-flash", + "modelRef": "qwen:qwen3.8-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "tool_use", + "agent", + "long_context", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.8-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "xhigh" + ], + "defaultEffort": "xhigh" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.8 Flash (Token Plan)", + "description": "百炼 Token Plan，通义千问 3.8 Flash 正式版，高速多模态推理", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true, + "thinkingMaxTokens": 262144, + "preserveThinkingDefault": true, + "supportsParallelToolCalls": false + }, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen3.8-flash" + } + }, + { + "apiModelId": "qwen3.7-max", + "modelRef": "qwen:qwen3.7-max", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.7-max", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.7 Max (Token Plan)", + "description": "百炼 Token Plan，通义千问 3.7 Max，1M 上下文，面向复杂推理与长程 Agent 任务", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen3.7-max" + } + }, + { + "apiModelId": "qwen3.7-plus", + "modelRef": "qwen:qwen3.7-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.7-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.7 Plus (Token Plan)", + "description": "百炼 Token Plan，通义千问 3.7 Plus，多模态智能体模型，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen3.7-plus" + } + }, + { + "apiModelId": "qwen3.6-plus", + "modelRef": "qwen:qwen3.6-plus", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.6-plus", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.6-plus", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.6 Plus (Token Plan)", + "description": "百炼 Token Plan，通义千问 3.6 Plus，1M 上下文，支持图像理解", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen3.6-plus" + } + }, + { + "apiModelId": "qwen3.6-flash", + "modelRef": "qwen:qwen3.6-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.6-flash", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "long_context", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#qwen3.6-flash", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Qwen3.6 Flash (Token Plan)", + "description": "百炼 Token Plan，通义千问 3.6 Flash，高性价比多模态模型，1M 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#qwen3.6-flash" + } + }, + { + "apiModelId": "kimi-k2.5", + "modelRef": "moonshot:kimi-k2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#kimi-k2.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#kimi-k2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.5 (Token Plan)", + "description": "百炼 Token Plan，Kimi K2.5，支持图像理解", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#kimi-k2.5" + } + }, + { + "apiModelId": "glm-5", + "modelRef": "zhipu:glm-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "contextWindow": { + "value": 202752, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#glm-5", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#glm-5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#glm-5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5 (Token Plan)", + "description": "百炼 Token Plan，智谱 GLM-5，200K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#glm-5" + } + }, + { + "apiModelId": "MiniMax-M2.5", + "modelRef": "minimax:MiniMax-M2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 196608, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#MiniMax-M2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5 (Token Plan)", + "description": "百炼 Token Plan，MiniMax M2.5，196K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#MiniMax-M2.5" + } + }, + { + "apiModelId": "glm-4.7", + "modelRef": "zhipu:glm-4.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/dashscope-token-plan.json#glm-4.7", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.7 (Token Plan)", + "description": "百炼 Token Plan，智谱 GLM-4.7，200K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/dashscope-token-plan.json#glm-4.7" + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/dashscope-token-plan.json" + } + }, + { + "id": "provider-infini-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-infini-coding-001", + "provider": "infini", + "brandGroup": "infini", + "label": "无问芯穹 Token (Coding) Plan", + "baseUrl": "https://cloud.infini-ai.com/maas/coding/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://cloud.infini-ai.com/console" + } + } + }, + "models": [ + { + "apiModelId": "deepseek-v3", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/infini-coding.json#deepseek-v3", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 128000, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/infini-coding.json#deepseek-v3", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 8192, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/infini-coding.json#deepseek-v3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/infini-coding.json#deepseek-v3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V3 (芯穹 Coding)", + "description": "无问芯穹 Coding Plan，DeepSeek V3 编程优化", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/infini-coding.json#deepseek-v3", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/infini-coding.json" + } + }, + { + "id": "provider-kwai-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-kwai-coding-001", + "provider": "kwai", + "brandGroup": "kwai", + "label": "快手 KwaiKAT Token (Coding) Plan", + "baseUrl": "https://wanqing.streamlakeapi.com/api/gateway/coding/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://console.streamlake.ai/wanqing/" + } + } + }, + "models": [ + { + "apiModelId": "kwai-coder", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/kwai-coding.json#kwai-coder", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "code" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/kwai-coding.json#kwai-coder", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "KwaiKAT Coder (Coding)", + "description": "快手 KwaiKAT Coding Plan，编程辅助模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/kwai-coding.json#kwai-coder", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/kwai-coding.json" + } + }, + { + "id": "provider-minimax-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-minimax-coding-001", + "provider": "minimax", + "brandGroup": "minimax", + "label": "MiniMax Token (Coding) Plan", + "baseUrl": "https://api.minimaxi.com/anthropic", + "apiFormat": "anthropic-messages", + "mediaBaseUrl": "https://api.minimaxi.com/v1", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "fast", + "reasoning", + "image_gen", + "tts", + "video_gen", + "music_gen" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "rest-api", + "endpoint": "https://api.minimaxi.com/v1/token_plan/remains", + "consoleUrl": "https://platform.minimaxi.com/" + } + } + }, + "models": [ + { + "apiModelId": "MiniMax-M3.1-Flash-Preview", + "modelRef": "minimax:MiniMax-M3.1-Flash-Preview", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "adaptiveThinking": true, + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "medium", + "high", + "xhigh", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M3.1 Flash Preview", + "description": "MiniMax M3.1 Flash Preview，目前仅通过 M Plan / MiniMax Code 提供", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#MiniMax-M3.1-Flash-Preview" + } + }, + { + "apiModelId": "MiniMax-M3", + "modelRef": "minimax:MiniMax-M3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#MiniMax-M3", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#MiniMax-M3", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "vision", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#MiniMax-M3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M3 (Token Plan)", + "description": "MiniMax M3，Token Plan 核心文本模型，1M 上下文，面向 Coding 与 Agent", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#MiniMax-M3" + } + }, + { + "apiModelId": "MiniMax-M2.7", + "modelRef": "minimax:MiniMax-M2.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7 (Token Plan)", + "description": "MiniMax 新一代旗舰模型，Token Plan 核心文本模型，5 小时滚动窗口计费。Starter 600次/5h，Plus 1500次/5h，Max 4500次/5h", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#MiniMax-M2.7" + } + }, + { + "apiModelId": "MiniMax-M2.7-highspeed", + "modelRef": "minimax:MiniMax-M2.7-highspeed", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7 高速版 (Token Plan)", + "description": "MiniMax M2.7 极速版，极速套餐专用。Plus-极速 1500次/5h，Max-极速 4500次/5h，Ultra-极速 30000次/5h", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#MiniMax-M2.7-highspeed" + } + }, + { + "apiModelId": "speech-2.8-hd", + "modelRef": "minimax:speech-2.8-hd", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "tts", + "multilingual", + "emotion_control", + "voice_clone" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#speech-2.8-hd", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Speech 2.8 HD (Token Plan)", + "description": "MiniMax 最新高清语音合成，Token Plan Plus 及以上可用。Plus 4000字符/天，Max 11000字符/天", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#speech-2.8-hd", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "speech-2.8-turbo", + "modelRef": "minimax:speech-2.8-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "tts", + "multilingual", + "emotion_control", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#speech-2.8-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Speech 2.8 Turbo (Token Plan)", + "description": "MiniMax 快速语音合成，Token Plan Plus 及以上可用", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#speech-2.8-turbo", + "adapterReadiness": { + "checkedAt": "2026-10-03", + "evidence": [ + { + "path": "packages/agent-service/src/speech/tts-adapter.ts", + "assertion": "recognize/synthesize通过resolveSpeechCompute解析model profile与adapter，非直接按Provider.apiFormat发送。" + }, + { + "path": "packages/agent-service/src/speech/speech-resolver.ts", + "assertion": "resolveProfile仅显式modelOrigin/speech或兼容表；无profile且无兼容命中立即fail missing_model_profile。" + }, + { + "path": "packages/speech-core/src/model-compatibility.ts", + "assertion": "逐精确ID匹配：tts-1/hd、whisper-1、mimo-v2.5tts/asr、cosyvoice-v2/v3前缀、fun-asr-realtime前缀；本行均未命中，无provider/local-whisper fallback。" + } + ], + "floorDecision": "没有实现/正式发布证据，不填猜测版本。候选展示保留，出站及自动路由禁止，适配验收后再解除。" + } + }, + "protocol": { + "state": "unsupported", + "reason": "missing-speech-profile-and-compatibility", + "apiFormat": "anthropic-messages", + "adapterId": "required-adapter:tts:minimax" + } + }, + { + "apiModelId": "image-01", + "modelRef": "minimax:image-01", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "image_generation", + "subject_reference" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#image-01", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Image 01 (Token Plan)", + "description": "MiniMax 文生图/图生图，Token Plan Plus 及以上可用。Plus 50张/天，Max 120张/天", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#image-01" + } + }, + { + "apiModelId": "image-01-live", + "modelRef": "minimax:image-01-live", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "image_generation", + "style_control", + "subject_reference" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#image-01-live", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Image 01 Live (Token Plan)", + "description": "MiniMax 画风控制文生图，Token Plan Plus 及以上可用", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#image-01-live" + } + }, + { + "apiModelId": "MiniMax-Hailuo-2.3", + "modelRef": "minimax:MiniMax-Hailuo-2.3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "text_to_video", + "image_to_video", + "camera_control" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#MiniMax-Hailuo-2.3", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 2.3 (Token Plan)", + "description": "海螺视频 2.3，文/图生视频，Token Plan Max 及以上可用。Max 2个/天", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#MiniMax-Hailuo-2.3" + } + }, + { + "apiModelId": "MiniMax-Hailuo-2.3-Fast", + "modelRef": "minimax:MiniMax-Hailuo-2.3-Fast", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "video_generation", + "image_to_video", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#MiniMax-Hailuo-2.3-Fast", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "海螺视频 2.3 快速版 (Token Plan)", + "description": "海螺视频快速版，Token Plan Max 及以上可用。Max 2个/天", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#MiniMax-Hailuo-2.3-Fast" + } + }, + { + "apiModelId": "music-2.6", + "modelRef": "minimax:music-2.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "music_generation", + "lyrics_input", + "instrumental" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/minimax-coding.json#music-2.6", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Music 2.6 (Token Plan)", + "description": "MiniMax AI 音乐生成，所有 Token Plan 套餐免费使用（100 首/天，≤5分钟/首）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/minimax-coding.json#music-2.6" + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/minimax-coding.json" + } + }, + { + "id": "provider-moonshot-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-moonshot-coding-001", + "provider": "moonshot", + "brandGroup": "moonshot", + "label": "Kimi Code", + "baseUrl": "https://api.kimi.com/coding/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://platform.moonshot.cn/" + } + } + }, + "models": [ + { + "apiModelId": "kimi-for-coding", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/moonshot-coding.json#kimi-for-coding", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/moonshot-coding.json#kimi-for-coding", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 32768, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/moonshot-coding.json#kimi-for-coding", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use", + "agent", + "long_context", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/moonshot-coding.json#kimi-for-coding", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi For Coding (Auto)", + "description": "Kimi Code 稳定路由 ID，自动映射到 Kimi Code 当前模型，262K 上下文", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/moonshot-coding.json#kimi-for-coding", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/moonshot-coding.json" + } + }, + { + "id": "provider-moorethread-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-moorethread-coding-001", + "provider": "moorethread", + "brandGroup": "moorethread", + "label": "摩尔线程 Token (Coding) Plan", + "baseUrl": "https://api.mthreads.com/v1", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://www.mthreads.com/console" + } + } + }, + "models": [ + { + "apiModelId": "mt-coder", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/moorethread-coding.json#mt-coder", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "code" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/moorethread-coding.json#mt-coder", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MT Coder (Coding)", + "description": "摩尔线程 Coding Plan，国产 GPU 推理", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/moorethread-coding.json#mt-coder", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/moorethread-coding.json" + } + }, + { + "id": "provider-tencent-token-001", + "category": "coding-plans", + "connection": { + "id": "provider-tencent-token-001", + "provider": "tencent", + "brandGroup": "tencent", + "label": "腾讯云 Token Plan", + "baseUrl": "https://api.lkeap.cloud.tencent.com/plan/v3", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "token-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://console.cloud.tencent.com/hunyuan" + } + } + }, + "models": [ + { + "apiModelId": "tc-code-latest", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#tc-code-latest", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://cloud.tencent.com/document/product/1823/130060", + "verification": "tool-use-per-official-plan-docs-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Auto (自动路由)", + "description": "腾讯 Token Plan 自动路由模型，后端自动选择最优模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#tc-code-latest", + "automaticRouting": false + } + }, + { + "apiModelId": "hunyuan-2.0-instruct", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-2.0-instruct", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "long_context", + "tool_use" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-2.0-instruct", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 2.0 Instruct", + "description": "腾讯混元 2.0 指令模型（Token Plan）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#hunyuan-2.0-instruct", + "automaticRouting": false + } + }, + { + "apiModelId": "hunyuan-2.0-thinking", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-2.0-thinking", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-2.0-thinking", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 2.0 Think", + "description": "腾讯混元 2.0 推理模型（Token Plan）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#hunyuan-2.0-thinking", + "automaticRouting": false + } + }, + { + "apiModelId": "hunyuan-t1", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-t1", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "deep_thinking" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-t1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 T1", + "description": "腾讯混元 T1 深度推理模型（Token Plan）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#hunyuan-t1", + "automaticRouting": false + } + }, + { + "apiModelId": "hunyuan-turbos", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-turbos", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#hunyuan-turbos", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "混元 Turbo S", + "description": "腾讯混元 Turbo S 高速模型（Token Plan）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#hunyuan-turbos", + "automaticRouting": false + } + }, + { + "apiModelId": "glm-5.1", + "modelRef": "zhipu:glm-5.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#glm-5.1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.1", + "description": "智谱 GLM-5.1（腾讯 Token Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#glm-5.1" + } + }, + { + "apiModelId": "glm-5", + "modelRef": "zhipu:glm-5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#glm-5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5", + "description": "智谱 GLM-5（腾讯 Token Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#glm-5" + } + }, + { + "apiModelId": "minimax-m2.7", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#minimax-m2.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://cloud.tencent.com/document/product/1823/130051", + "verification": "tool-use-officially-verified-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7", + "description": "MiniMax M2.7（腾讯 Token Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#minimax-m2.7", + "automaticRouting": false + } + }, + { + "apiModelId": "minimax-m2.5", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#minimax-m2.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#minimax-m2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5", + "description": "MiniMax M2.5（腾讯 Token Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#minimax-m2.5", + "automaticRouting": false + } + }, + { + "apiModelId": "kimi-k2.5", + "modelRef": "moonshot:kimi-k2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/tencent-token.json#kimi-k2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.5", + "description": "Kimi K2.5（腾讯 Token Plan 渠道）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/tencent-token.json#kimi-k2.5" + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/tencent-token.json" + } + }, + { + "id": "provider-volcengine-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-volcengine-coding-001", + "provider": "volcengine", + "brandGroup": "volcengine", + "label": "火山引擎 Token (Coding) Plan", + "baseUrl": "https://ark.cn-beijing.volces.com/api/coding/v3", + "apiFormat": "openai-completions", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "response-header", + "headerKeys": { + "remaining": "x-ratelimit-remaining", + "limit": "x-ratelimit-limit", + "reset": "x-ratelimit-reset" + }, + "consoleUrl": "https://console.volcengine.com/ark" + }, + "modelIdOverride": "ark-code-latest" + } + }, + "models": [ + { + "apiModelId": "doubao-seed-2.1-pro", + "modelRef": "volcengine:doubao-seed-2.1-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.1-pro", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.1-pro", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.1 Pro", + "description": "火山方舟 Coding Plan，豆包 Seed 2.1 Pro，生产级 Coding 与 Agent 模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.1-pro" + } + }, + { + "apiModelId": "doubao-seed-2.1-turbo", + "modelRef": "volcengine:doubao-seed-2.1-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "maxOutputTokens": { + "value": 262144, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "vision", + "tool_use", + "agent", + "fast", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.1-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.1 Turbo", + "description": "火山方舟 Coding Plan，豆包 Seed 2.1 Turbo，高吞吐 Coding 与 Agent 模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.1-turbo" + } + }, + { + "apiModelId": "ark-code-latest", + "modelRef": null, + "binding": { + "kind": "dynamic", + "reason": "动态目录/套餐滚动ID；必须按真实返回目录识别，不静态绑猜测原厂，未识别期间保留access-local。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#ark-code-latest", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "tool_use" + ], + "reason": "保留原接入标签；2026-10-05 按官网补 tool_use，其余标签未重新核验。", + "evidence": "https://docs.volcengine.com/docs/ark/coding-plan-personal-get-started?lang=zh", + "verification": "tool-use-per-official-plan-docs-other-tags-inherited" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Ark Code Latest (Auto)", + "description": "自动路由最优编程模型（Doubao Seed Code / GLM-4.7 / DeepSeek-V3.2 / Kimi-K2.5 等）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#ark-code-latest", + "automaticRouting": false + } + }, + { + "apiModelId": "doubao-seed-2.0-code", + "modelRef": "volcengine:doubao-seed-2.0-code", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.0-code", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.0 Code", + "description": "火山方舟 Coding Plan 中的豆包 Seed 2.0 编程模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.0-code" + } + }, + { + "apiModelId": "doubao-seed-2.0-pro", + "modelRef": "volcengine:doubao-seed-2.0-pro", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "long_context", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.0-pro", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.0 Pro", + "description": "火山方舟 Coding Plan 中的豆包 Seed 2.0 旗舰模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.0-pro" + } + }, + { + "apiModelId": "doubao-seed-2.0-lite", + "modelRef": "volcengine:doubao-seed-2.0-lite", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "long_context", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.0-lite", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed 2.0 Lite", + "description": "火山方舟 Coding Plan 中的豆包 Seed 2.0 均衡模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#doubao-seed-2.0-lite" + } + }, + { + "apiModelId": "doubao-seed-code", + "modelRef": "volcengine:doubao-seed-code", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#doubao-seed-code", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "豆包 Seed Code", + "description": "火山方舟 Coding Plan 中的豆包编程模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#doubao-seed-code" + } + }, + { + "apiModelId": "glm-4.7", + "modelRef": "zhipu:glm-4.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#glm-4.7", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.7", + "description": "火山方舟 Coding Plan 中的智谱 GLM-4.7 模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#glm-4.7" + } + }, + { + "apiModelId": "deepseek-v3.2", + "modelRef": "volcengine:deepseek-v3.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#deepseek-v3.2", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "DeepSeek V3.2", + "description": "火山方舟 Coding Plan 中的 DeepSeek V3.2 模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#deepseek-v3.2" + } + }, + { + "apiModelId": "kimi-k2.5", + "modelRef": "moonshot:kimi-k2.5", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#kimi-k2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.5", + "description": "火山方舟 Coding Plan 中的 Kimi K2.5 模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#kimi-k2.5" + } + }, + { + "apiModelId": "glm-5.1", + "modelRef": "zhipu:glm-5.1", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "long_context", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#glm-5.1", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.1", + "description": "火山方舟 Coding Plan 中的智谱 GLM-5.1 旗舰模型（2026.4 新增）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#glm-5.1" + } + }, + { + "apiModelId": "kimi-k2.6", + "modelRef": "moonshot:kimi-k2.6", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "agent", + "long_context", + "vision" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#kimi-k2.6", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "Kimi K2.6", + "description": "火山方舟 Coding Plan 中的 Kimi K2.6 模型（2026.4 新增）", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#kimi-k2.6" + } + }, + { + "apiModelId": "minimax-m2.5", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#minimax-m2.5", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "agent" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#minimax-m2.5", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.5", + "description": "火山方舟 Coding Plan 中的 MiniMax M2.5 模型", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#minimax-m2.5", + "automaticRouting": false + } + }, + { + "apiModelId": "minimax-m2.7", + "modelRef": null, + "binding": { + "kind": "access-local", + "reason": "精确身份未证实；保留接入本地字段，禁止normalized/stripped/family制造原厂绑定。" + }, + "overrides": { + "serviceType": { + "value": [ + "chat" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#minimax-m2.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "code", + "reasoning", + "chat", + "agent", + "long_context" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/volcengine-coding.json#minimax-m2.7", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "MiniMax M2.7", + "description": "火山方舟 Coding Plan 中的 MiniMax M2.7 模型（2026.4 新增），230B 总参/10B 激活", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/volcengine-coding.json#minimax-m2.7", + "automaticRouting": false + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/volcengine-coding.json" + } + }, + { + "id": "provider-zhipu-coding-001", + "category": "coding-plans", + "connection": { + "id": "provider-zhipu-coding-001", + "provider": "zhipu", + "brandGroup": "zhipu", + "label": "智谱 Token (Coding) Plan", + "baseUrl": "https://api.z.ai/api/anthropic", + "apiFormat": "anthropic-messages", + "apiKeyRef": "", + "apiKeyVerified": false, + "enabled": false, + "status": "unconfigured", + "accessMode": "coding-plan", + "services": [ + "chat", + "reasoning" + ], + "codingPlan": { + "quotas": {}, + "usageTracking": { + "method": "none", + "consoleUrl": "https://open.bigmodel.cn/console/overview" + } + }, + "tombstones": [ + "glm-5.1", + "glm-5", + "glm-4.5-air" + ] + }, + "models": [ + { + "apiModelId": "glm-5.3", + "modelRef": "zhipu:glm-5.3", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": {}, + "request": { + "adaptiveThinking": true, + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [ + { + "sourceKey": "extra.pricingNotes", + "amount": "国内官方标准价，人民币/百万 tokens；Flash 限时优惠以控制台实际结算为准。", + "currency": "unknown", + "unit": "annotation", + "verification": "inherited-not-reverified" + } + ], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.3", + "description": "智谱 GLM-5.3 大推理模型，面向复杂软件工程与长程 Agent 任务，100 万上下文；强制思考，无法关闭", + "source": "preset" + }, + "recordedExtra": { + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/zhipu-coding.json#glm-5.3" + } + }, + { + "apiModelId": "glm-5.3-flash", + "modelRef": "zhipu:glm-5.3-flash", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "capabilities": { + "value": [ + "chat", + "reasoning", + "deep_thinking", + "code", + "vision", + "image_understanding", + "video_understanding", + "long_context", + "tool_use", + "agent", + "multilingual", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5.3-flash", + "verification": "inherited-not-reverified" + } + }, + "request": { + "adaptiveThinking": true, + "constraints": { + "thinkingOnly": true + }, + "reasoning": { + "supportedEfforts": [ + "low", + "high", + "max" + ], + "defaultEffort": "max" + } + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.3-Flash (Coding)", + "description": "智谱 GLM-5.3-Flash Coding Plan 版本，原生多模态高效模型，1M 上下文，128K 最大输出，强制开启思考", + "source": "preset" + }, + "recordedExtra": { + "supportsThinking": true, + "thinkingDefault": true + }, + "sourceRecord": "compute/coding-plans/zhipu-coding.json#glm-5.3-flash" + } + }, + { + "apiModelId": "glm-5.2", + "modelRef": "zhipu:glm-5.2", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "contextWindow": { + "value": 1000000, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5.2", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "math", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5.2", + "verification": "inherited-not-reverified" + } + }, + "request": { + "reasoning": { + "supportedEfforts": [ + "max", + "xhigh", + "high", + "medium", + "low", + "minimal", + "none" + ], + "defaultEffort": "max" + }, + "upstreamModelId": "glm-5.2[1m]" + }, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5.2 (Coding)", + "description": "智谱 GLM-5.2 Coding Plan 版本，官方最新可用模型，1M 上下文，128K 最大输出", + "source": "preset" + }, + "recordedExtra": { + "codingPlanLongContextModelId": "glm-5.2[1m]" + }, + "sourceRecord": "compute/coding-plans/zhipu-coding.json#glm-5.2" + } + }, + { + "apiModelId": "glm-5-turbo", + "modelRef": "zhipu:glm-5-turbo", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5-turbo", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5-turbo", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "math", + "multilingual", + "fast" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-5-turbo", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-5-Turbo (Coding)", + "description": "智谱 GLM-5-Turbo Coding Plan 版本，官方可用高阶模型，200K 上下文，128K 最大输出", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/zhipu-coding.json#glm-5-turbo" + } + }, + { + "apiModelId": "glm-4.7", + "modelRef": "zhipu:glm-4.7", + "binding": { + "kind": "exact", + "reason": "exact-id-or-declared-alias" + }, + "overrides": { + "serviceType": { + "value": [ + "chat", + "reasoning" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "maxOutputTokens": { + "value": 131072, + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-4.7", + "verification": "inherited-not-reverified" + }, + "capabilities": { + "value": [ + "chat", + "reasoning", + "code", + "deep_thinking", + "long_context", + "tool_use", + "agent", + "multilingual" + ], + "reason": "preserved-access-contract", + "evidence": "compute/coding-plans/zhipu-coding.json#glm-4.7", + "verification": "inherited-not-reverified" + } + }, + "request": {}, + "billing": [], + "lifecycle": { + "status": "unknown" + }, + "metadata": { + "attributes": { + "displayName": "GLM-4.7 (Coding)", + "description": "智谱 GLM-4.7 Coding Plan 版本，官方可用模型，200K 上下文，128K 最大输出", + "source": "preset" + }, + "recordedExtra": {}, + "sourceRecord": "compute/coding-plans/zhipu-coding.json#glm-4.7" + } + } + ], + "metadata": { + "sourceFile": "compute/coding-plans/zhipu-coding.json" + } + } + ], + "retirements": [ + { + "id": "config-center-lingyiwanwu-access-retirement", + "kind": "product-access-retired", + "scope": "managed-provider", + "providerId": "provider-lingyiwanwu-001", + "origin": { + "exactHost": "api.lingyiwanwu.com", + "pathPrefix": "/v1" + }, + "effectiveAt": "2026-08-28T06:52:05Z", + "reason": "config-center #84 已移除零一万物预置接入；存量受管配置不再作为官方可调用接入。", + "evidence": { + "kind": "product-repository", + "url": "https://github.com/desirecore/config-center/commit/a998e18e6c4bdcc161f0baf82cbf534cc9d53627", + "checkedAt": "2026-10-03" + }, + "preserveCredentials": true, + "vendorDomainShutdownConfirmed": false + } + ], + "metadata": { + "migrationEpoch": 1, + "authority": "catalog/v2/catalog.json", + "verification": "migration-preserves-existing-evidence" + }, + "digest": "8f0c6ee356bb44e60fb17d1dca04b012d47cae9cea9fd7890f874894ee74cdb5" +} diff --git a/manifest.json b/manifest.json new file mode 100644 index 0000000..ce7b7ce --- /dev/null +++ b/manifest.json @@ -0,0 +1,6 @@ +{ + "version": "1.0.0", + "presetDataVersion": 131, + "updatedAt": "2026-10-13", + "description": "DesireCore 官方配置中心" +} diff --git a/runtimes/recommended.json b/runtimes/recommended.json new file mode 100644 index 0000000..1537af9 --- /dev/null +++ b/runtimes/recommended.json @@ -0,0 +1,44 @@ +{ + "python": { + "name": "3.13", + "version": "3.13.9", + "distTag": "20251014", + "archives": { + "darwin-arm64": "20251014/cpython-3.13.9+20251014-aarch64-apple-darwin-install_only_stripped.tar.gz", + "darwin-x64": "20251014/cpython-3.13.9+20251014-x86_64-apple-darwin-install_only_stripped.tar.gz", + "win32-x64": "20251014/cpython-3.13.9+20251014-x86_64-pc-windows-msvc-install_only_stripped.tar.gz", + "win32-arm64": "20251014/cpython-3.13.9+20251014-aarch64-pc-windows-msvc-install_only_stripped.tar.gz", + "linux-x64": "20251014/cpython-3.13.9+20251014-x86_64-unknown-linux-gnu-install_only_stripped.tar.gz", + "linux-arm64": "20251014/cpython-3.13.9+20251014-aarch64-unknown-linux-gnu-install_only_stripped.tar.gz" + }, + "sha256": { + "darwin-arm64": "52721745b0fa3196e4d0381fa5c06dda1d54343b90d49d90c3bba52d1171bd98", + "darwin-x64": "7c33b153a69c6255e6f2659cf39738f316b03969d6230d7bc47c73b7fde9a0d4", + "win32-x64": "76e0a9749c4deeb975a4b6b36d54be4e43f0c2a4c654bedab5d2e4d62dbc3006", + "win32-arm64": "419cce7a099d21389fca32a9fd4f7364fc61b91a760369c803de57c6ee1ddb99", + "linux-x64": "c0ccda275948c79996e46993c2c5476ff5cca606dee530f1dea712179131b348", + "linux-arm64": "baa3d107d17e4328448e30c3c9c83cca0eb41ca7a37c10982e14d46a5c3db07d" + } + }, + "node": { + "version": "24.21.0", + "ltsCodename": "Krypton", + "archives": { + "darwin-arm64": "v24.21.0/node-v24.21.0-darwin-arm64.tar.gz", + "darwin-x64": "v24.21.0/node-v24.21.0-darwin-x64.tar.gz", + "win32-x64": "v24.21.0/node-v24.21.0-win-x64.zip", + "win32-arm64": "v24.21.0/node-v24.21.0-win-arm64.zip", + "linux-x64": "v24.21.0/node-v24.21.0-linux-x64.tar.gz", + "linux-arm64": "v24.21.0/node-v24.21.0-linux-arm64.tar.gz" + }, + "sha256": { + "darwin-arm64": "bed7eea5325e1108f32ce5228ddd6a5f0f08a499ee42aa7442aea583702f6057", + "darwin-x64": "1462cb3b3046b815cf8ea436d3da450ec1a9f11dac7e5a46b0ada5305d7e8097", + "win32-x64": "158f7685b44de51f6c0df1d153526cbcd3e1bc739a8dfc607721cef75de9e541", + "win32-arm64": "8779b1bde1d39f8d420e3b57aa657b39891af434d3de44a919044cec06785921", + "linux-x64": "6e1db87ef58b8819e5d5402eff1536491b18edd8eb7bee5ef7897876e88dc5ff", + "linux-arm64": "724282c3b43aec998aa9527380465b45d229e021b58035f5f4f63095eabfe5d5" + } + }, + "minClientVersion": "10.0.178" +} diff --git a/runtimes/versions-fallback.json b/runtimes/versions-fallback.json new file mode 100644 index 0000000..4cdfdc0 --- /dev/null +++ b/runtimes/versions-fallback.json @@ -0,0 +1,1437 @@ +{ + "generatedAt": "2026-10-03T10:24:20.059Z", + "node": [ + { + "version": "v26.10.0", + "date": "2026-09-21", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "linux-x64-musl", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.19.1", + "v8": "14.6.202.34", + "uv": "1.52.1", + "zlib": "1.3.2.1-motley", + "openssl": "3.5.8", + "modules": "147", + "lts": false, + "security": false + }, + { + "version": "v26.9.0", + "date": "2026-09-16", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "linux-x64-musl", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.19.1", + "v8": "14.6.202.34", + "uv": "1.52.1", + "zlib": "1.3.2.1-motley", + "openssl": "3.5.8", + "modules": "147", + "lts": false, + "security": false + }, + { + "version": "v26.8.2", + "date": "2026-09-09", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "linux-x64-musl", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.19.1", + "v8": "14.6.202.34", + "uv": "1.52.1", + "zlib": "1.3.2.1-motley", + "openssl": "3.5.8", + "modules": "147", + "lts": false, + "security": false + }, + { + "version": "v25.9.0", + "date": "2026-03-31", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.12.1", + "v8": "14.1.146.11", + "uv": "1.51.0", + "zlib": "1.3.1", + "openssl": "3.5.5", + "modules": "141", + "lts": false, + "security": false + }, + { + "version": "v25.8.2", + "date": "2026-03-24", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.11.1", + "v8": "14.1.146.11", + "uv": "1.51.0", + "zlib": "1.3.1", + "openssl": "3.5.5", + "modules": "141", + "lts": false, + "security": true + }, + { + "version": "v25.8.1", + "date": "2026-03-10", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.11.0", + "v8": "14.1.146.11", + "uv": "1.51.0", + "zlib": "1.3.1", + "openssl": "3.5.5", + "modules": "141", + "lts": false, + "security": false + }, + { + "version": "v24.21.0", + "date": "2026-09-07", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "linux-x64-musl", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.19.0", + "v8": "13.6.233.17", + "uv": "1.52.1", + "zlib": "1.3.2.1-motley", + "openssl": "3.5.8", + "modules": "137", + "lts": "Krypton", + "security": false + }, + { + "version": "v24.20.0", + "date": "2026-08-26", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "linux-x64-musl", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.19.0", + "v8": "13.6.233.17", + "uv": "1.52.1", + "zlib": "1.3.2.1-motley", + "openssl": "3.5.7", + "modules": "137", + "lts": "Krypton", + "security": false + }, + { + "version": "v24.19.0", + "date": "2026-08-03", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "11.17.0", + "v8": "13.6.233.17", + "uv": "1.52.1", + "zlib": "1.3.2.1-motley", + "openssl": "3.5.7", + "modules": "137", + "lts": "Krypton", + "security": false + }, + { + "version": "v23.11.1", + "date": "2025-05-14", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "10.9.2", + "v8": "12.9.202.28", + "uv": "1.50.0", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.16", + "modules": "131", + "lts": false, + "security": true + }, + { + "version": "v23.11.0", + "date": "2025-04-01", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "10.9.2", + "v8": "12.9.202.28", + "uv": "1.50.0", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.16", + "modules": "131", + "lts": false, + "security": false + }, + { + "version": "v23.10.0", + "date": "2025-03-13", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip" + ], + "npm": "10.9.2", + "v8": "12.9.202.28", + "uv": "1.50.0", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.16", + "modules": "131", + "lts": false, + "security": false + }, + { + "version": "v22.23.3", + "date": "2026-09-23", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "linux-x64-musl", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.9.9", + "v8": "12.4.254.21", + "uv": "1.51.0", + "zlib": "1.3.1", + "openssl": "3.5.8", + "modules": "127", + "lts": "Jod", + "security": false + }, + { + "version": "v22.23.2", + "date": "2026-07-28", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.9.8", + "v8": "12.4.254.21", + "uv": "1.51.0", + "zlib": "1.3.1", + "openssl": "3.5.7", + "modules": "127", + "lts": "Jod", + "security": true + }, + { + "version": "v22.23.1", + "date": "2026-06-22", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.9.8", + "v8": "12.4.254.21", + "uv": "1.51.0", + "zlib": "1.3.1", + "openssl": "3.5.7", + "modules": "127", + "lts": "Jod", + "security": false + }, + { + "version": "v21.7.3", + "date": "2024-04-10", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.5.0", + "v8": "11.8.172.17", + "uv": "1.48.0", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.13+quic", + "modules": "120", + "lts": false, + "security": true + }, + { + "version": "v21.7.2", + "date": "2024-04-03", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.5.0", + "v8": "11.8.172.17", + "uv": "1.48.0", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.13+quic", + "modules": "120", + "lts": false, + "security": true + }, + { + "version": "v21.7.1", + "date": "2024-03-08", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.5.0", + "v8": "11.8.172.17", + "uv": "1.48.0", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.13+quic", + "modules": "120", + "lts": false, + "security": false + }, + { + "version": "v20.20.2", + "date": "2026-03-24", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.8.2", + "v8": "11.3.244.8", + "uv": "1.46.0", + "zlib": "1.3.1", + "openssl": "3.0.19", + "modules": "115", + "lts": "Iron", + "security": true + }, + { + "version": "v20.20.1", + "date": "2026-03-04", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.8.2", + "v8": "11.3.244.8", + "uv": "1.46.0", + "zlib": "1.3.1", + "openssl": "3.0.19", + "modules": "115", + "lts": "Iron", + "security": false + }, + { + "version": "v20.20.0", + "date": "2026-01-12", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.8.2", + "v8": "11.3.244.8", + "uv": "1.46.0", + "zlib": "1.3.1", + "openssl": "3.0.17", + "modules": "115", + "lts": "Iron", + "security": true + }, + { + "version": "v19.9.0", + "date": "2023-04-10", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-arm64-7z", + "win-arm64-zip", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "9.6.3", + "v8": "10.8.168.25", + "uv": "1.44.2", + "zlib": "1.2.13", + "openssl": "3.0.8+quic", + "modules": "111", + "lts": false, + "security": false + }, + { + "version": "v19.8.1", + "date": "2023-03-15", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "9.5.1", + "v8": "10.8.168.25", + "uv": "1.44.2", + "zlib": "1.2.13", + "openssl": "3.0.8+quic", + "modules": "111", + "lts": false, + "security": false + }, + { + "version": "v19.8.0", + "date": "2023-03-14", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "9.5.1", + "v8": "10.8.168.25", + "uv": "1.44.2", + "zlib": "1.2.13", + "openssl": "3.0.8+quic", + "modules": "111", + "lts": false, + "security": false + }, + { + "version": "v18.20.8", + "date": "2025-03-27", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.8.2", + "v8": "10.2.154.26", + "uv": "1.44.2", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.16", + "modules": "108", + "lts": "Hydrogen", + "security": false + }, + { + "version": "v18.20.7", + "date": "2025-02-20", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.8.2", + "v8": "10.2.154.26", + "uv": "1.44.2", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.15+quic", + "modules": "108", + "lts": "Hydrogen", + "security": false + }, + { + "version": "v18.20.6", + "date": "2025-01-21", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "10.8.2", + "v8": "10.2.154.26", + "uv": "1.44.2", + "zlib": "1.3.0.1-motley", + "openssl": "3.0.15+quic", + "modules": "108", + "lts": "Hydrogen", + "security": true + }, + { + "version": "v17.9.1", + "date": "2022-06-01", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "8.11.0", + "v8": "9.6.180.15", + "uv": "1.43.0", + "zlib": "1.2.11", + "openssl": "3.0.3+quic", + "modules": "102", + "lts": false, + "security": false + }, + { + "version": "v17.9.0", + "date": "2022-04-07", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "8.5.5", + "v8": "9.6.180.15", + "uv": "1.43.0", + "zlib": "1.2.11", + "openssl": "3.0.2+quic", + "modules": "102", + "lts": false, + "security": false + }, + { + "version": "v17.8.0", + "date": "2022-03-22", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "8.5.5", + "v8": "9.6.180.15", + "uv": "1.43.0", + "zlib": "1.2.11", + "openssl": "3.0.2+quic", + "modules": "102", + "lts": false, + "security": false + }, + { + "version": "v16.20.2", + "date": "2023-08-08", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "8.19.4", + "v8": "9.4.146.26", + "uv": "1.43.0", + "zlib": "1.2.11", + "openssl": "1.1.1v+quic", + "modules": "93", + "lts": "Gallium", + "security": true + }, + { + "version": "v16.20.1", + "date": "2023-06-20", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "8.19.4", + "v8": "9.4.146.26", + "uv": "1.43.0", + "zlib": "1.2.11", + "openssl": "1.1.1u+quic", + "modules": "93", + "lts": "Gallium", + "security": true + }, + { + "version": "v16.20.0", + "date": "2023-03-28", + "files": [ + "aix-ppc64", + "headers", + "linux-arm64", + "linux-armv7l", + "linux-ppc64le", + "linux-s390x", + "linux-x64", + "osx-arm64-tar", + "osx-x64-pkg", + "osx-x64-tar", + "src", + "win-x64-7z", + "win-x64-exe", + "win-x64-msi", + "win-x64-zip", + "win-x86-7z", + "win-x86-exe", + "win-x86-msi", + "win-x86-zip" + ], + "npm": "8.19.4", + "v8": "9.4.146.26", + "uv": "1.43.0", + "zlib": "1.2.11", + "openssl": "1.1.1t+quic", + "modules": "93", + "lts": "Gallium", + "security": false + } + ], + "python": [ + { + "name": "3.7", + "version": "3.7.9" + }, + { + "name": "3.8", + "version": "3.8.20" + }, + { + "name": "3.9", + "version": "3.9.24" + }, + { + "name": "3.10", + "version": "3.10.19" + }, + { + "name": "3.11", + "version": "3.11.14" + }, + { + "name": "3.12", + "version": "3.12.12" + }, + { + "name": "3.13", + "version": "3.13.9" + }, + { + "name": "3.14", + "version": "3.14.0" + }, + { + "name": "pypy2.7", + "version": "7.3.20" + }, + { + "name": "pypy3.9", + "version": "7.3.16" + }, + { + "name": "pypy3.10", + "version": "7.3.19" + }, + { + "name": "pypy3.11", + "version": "7.3.20" + } + ], + "pkgManagers": { + "npm": [ + { + "tag_name": "v12.2.0", + "created_at": "2026-09-30T16:47:29Z" + }, + { + "tag_name": "v12.1.0", + "created_at": "2026-09-22T20:21:08Z" + }, + { + "tag_name": "v11.20.0", + "created_at": "2026-09-22T21:34:44Z" + }, + { + "tag_name": "v11.19.1", + "created_at": "2026-08-26T21:48:47Z" + }, + { + "tag_name": "v12.0.2", + "created_at": "2026-07-29T21:11:46Z" + }, + { + "tag_name": "v11.19.0", + "created_at": "2026-07-29T20:57:30Z" + }, + { + "tag_name": "v10.9.9", + "created_at": "2026-07-29T18:02:54Z" + }, + { + "tag_name": "v12.0.1", + "created_at": "2026-07-10T17:18:50Z" + }, + { + "tag_name": "v12.0.0", + "created_at": "2026-07-08T21:21:30Z" + }, + { + "tag_name": "v11.18.0", + "created_at": "2026-06-29T16:48:15Z" + }, + { + "tag_name": "v11.17.0", + "created_at": "2026-06-11T18:48:56Z" + }, + { + "tag_name": "v11.16.0", + "created_at": "2026-05-27T20:52:19Z" + }, + { + "tag_name": "v11.15.0", + "created_at": "2026-05-20T20:41:46Z" + } + ], + "pnpm": [ + { + "tag_name": "v12.9.0", + "created_at": "2026-10-02T22:28:05Z" + }, + { + "tag_name": "v12.8.2", + "created_at": "2026-09-30T01:08:47Z" + }, + { + "tag_name": "v11.28.3", + "created_at": "2026-09-30T15:28:07Z" + }, + { + "tag_name": "v12.8.1", + "created_at": "2026-09-28T17:40:12Z" + }, + { + "tag_name": "v12.8.0", + "created_at": "2026-09-28T07:35:50Z" + }, + { + "tag_name": "v11.28.2", + "created_at": "2026-09-28T17:40:47Z" + }, + { + "tag_name": "v11.28.1", + "created_at": "2026-09-28T07:31:11Z" + }, + { + "tag_name": "v10.34.6", + "created_at": "2026-09-29T15:44:40Z" + }, + { + "tag_name": "v12.7.0", + "created_at": "2026-09-25T10:39:37Z" + }, + { + "tag_name": "v11.28.0", + "created_at": "2026-09-25T10:39:54Z" + }, + { + "tag_name": "v12.6.0", + "created_at": "2026-09-22T17:08:42Z" + }, + { + "tag_name": "v11.27.1", + "created_at": "2026-09-20T21:38:41Z" + }, + { + "tag_name": "v12.5.1", + "created_at": "2026-09-18T21:48:32Z" + }, + { + "tag_name": "v12.5.0", + "created_at": "2026-09-18T17:02:57Z" + }, + { + "tag_name": "v12.4.2", + "created_at": "2026-09-15T10:49:15Z" + }, + { + "tag_name": "v11.27.0", + "created_at": "2026-09-12T21:22:42Z" + }, + { + "tag_name": "v12.4.1", + "created_at": "2026-09-10T17:00:15Z" + }, + { + "tag_name": "v12.4.0", + "created_at": "2026-09-08T11:27:26Z" + }, + { + "tag_name": "v11.26.0", + "created_at": "2026-09-06T23:50:11Z" + }, + { + "tag_name": "v12.3.4", + "created_at": "2026-09-04T14:23:18Z" + }, + { + "tag_name": "v12.3.3", + "created_at": "2026-09-04T14:23:32Z" + }, + { + "tag_name": "v12.3.2", + "created_at": "2026-09-04T02:07:17Z" + }, + { + "tag_name": "v12.3.1", + "created_at": "2026-09-03T00:40:37Z" + }, + { + "tag_name": "v12.3.0", + "created_at": "2026-09-02T16:28:57Z" + }, + { + "tag_name": "v12.2.1", + "created_at": "2026-09-01T00:57:45Z" + }, + { + "tag_name": "v12.2.0", + "created_at": "2026-09-01T00:01:41Z" + }, + { + "tag_name": "v12.1.0", + "created_at": "2026-08-29T14:19:52Z" + }, + { + "tag_name": "v11.25.0", + "created_at": "2026-08-29T14:20:10Z" + }, + { + "tag_name": "v12.0.0", + "created_at": "2026-08-26T15:13:51Z" + }, + { + "tag_name": "v11.24.0", + "created_at": "2026-08-24T19:10:25Z" + } + ], + "yarn": [ + { + "tag_name": "v4.18.1", + "created_at": "2026-09-24T20:40:35Z" + }, + { + "tag_name": "v4.18.0", + "created_at": "2026-07-29T07:56:33Z" + }, + { + "tag_name": "v4.17.1", + "created_at": "2026-07-08T16:48:56Z" + }, + { + "tag_name": "v4.17.0", + "created_at": "2026-06-15T08:28:48Z" + }, + { + "tag_name": "v4.16.0", + "created_at": "2026-06-02T07:11:32Z" + }, + { + "tag_name": "v4.15.0", + "created_at": "2026-05-19T22:20:46Z" + }, + { + "tag_name": "v4.14.1", + "created_at": "2026-04-17T06:00:10Z" + }, + { + "tag_name": "v4.14.0", + "created_at": "2026-04-16T15:45:42Z" + }, + { + "tag_name": "v4.13.0", + "created_at": "2026-03-03T02:12:57Z" + }, + { + "tag_name": "v4.12.0", + "created_at": "2025-11-23T11:09:03Z" + }, + { + "tag_name": "v4.11.0", + "created_at": "2025-11-07T13:52:43Z" + }, + { + "tag_name": "v4.10.3", + "created_at": "2025-09-23T09:42:00Z" + }, + { + "tag_name": "v4.10.2", + "created_at": "2025-09-18T17:57:07Z" + }, + { + "tag_name": "v4.10.1", + "created_at": "2025-09-18T12:53:06Z" + }, + { + "tag_name": "v4.10.0", + "created_at": "2025-09-18T09:18:24Z" + }, + { + "tag_name": "v4.9.4", + "created_at": "2025-08-26T07:12:31Z" + }, + { + "tag_name": "v4.9.3", + "created_at": "2025-08-20T12:15:54Z" + }, + { + "tag_name": "v4.9.2", + "created_at": "2025-06-03T20:25:28Z" + }, + { + "tag_name": "v4.9.1", + "created_at": "2025-04-11T09:04:22Z" + }, + { + "tag_name": "v4.9.0", + "created_at": "2025-04-09T14:12:16Z" + }, + { + "tag_name": "v4.8.1", + "created_at": "2025-03-30T08:45:27Z" + }, + { + "tag_name": "v4.8.0", + "created_at": "2025-03-28T10:35:51Z" + }, + { + "tag_name": "v4.7.0", + "created_at": "2025-03-01T15:19:32Z" + }, + { + "tag_name": "v4.6.0", + "created_at": "2024-12-29T12:34:29Z" + }, + { + "tag_name": "v3.8.7", + "created_at": "2024-12-04T00:54:22Z" + }, + { + "tag_name": "v4.5.3", + "created_at": "2024-11-25T13:51:20Z" + }, + { + "tag_name": "v4.5.2", + "created_at": "2024-11-22T00:33:09Z" + }, + { + "tag_name": "v4.5.1", + "created_at": "2024-10-18T16:19:19Z" + }, + { + "tag_name": "v3.8.6", + "created_at": "2024-10-18T17:10:37Z" + }, + { + "tag_name": "v4.5.0", + "created_at": "2024-09-14T12:52:51Z" + } + ] + } +} diff --git a/schemas/model-spec.schema.json b/schemas/model-spec.schema.json new file mode 100644 index 0000000..422ca02 --- /dev/null +++ b/schemas/model-spec.schema.json @@ -0,0 +1,311 @@ +{ + "$schema": "http://json-schema.org/draft-07/schema#", + "$id": "https://desirecore.net/schemas/config-center/model-spec.schema.json", + "title": "ModelSpecFile", + "description": "模型规格库（compute/model-specs/*.json）数据契约。\n\n用途：维护模型的唯一规格主数据：内在参数（上下文窗口、最大输出、能力标签、serviceType、默认采样参数等）以及智能路由策略（量级、排序、Agent 可选性和标准化 reasoning 合同）。同一事实不得再写入独立 smart-routing 目录。\n\nProvider 文件只描述接入面、可用性和明确的接入面覆盖；历史 Provider 条目中已有的规格字段为旧客户端兼容副本，不能覆盖本文件。\n\n严禁写入价格字段（inputPrice/outputPrice/cache 等）——价签真相源永远是 NewAPI pricing。", + "type": "object", + "required": ["specs"], + "properties": { + "description": { + "type": "string", + "description": "文件说明（如所属厂商、收录范围）" + }, + "specs": { + "type": "array", + "description": "模型规格条目列表", + "items": { "$ref": "#/definitions/modelSpec" } + } + }, + "additionalProperties": false, + "definitions": { + "modelOrigin": { + "type": "object", + "required": ["vendor", "canonicalModelId"], + "properties": { + "vendor": { + "type": "string", + "minLength": 1, + "pattern": "^[a-z0-9]+(?:-[a-z0-9]+)*$", + "description": "规范化原厂商 ID，如 xiaomi、openai、alibaba、minimax；不得填写当前算力 Provider ID。" + }, + "canonicalModelId": { + "type": "string", + "minLength": 1, + "description": "原厂模型的稳定识别 ID，用于适配器匹配、诊断和日志；不替代实际路由使用的 modelName/apiModelId。" + }, + "family": { + "type": "string", + "minLength": 1, + "description": "跨小版本稳定的模型家族，如 mimo-asr；同一协议下的后续模型版本可复用该家族语义。" + } + }, + "additionalProperties": false + }, + "speechVoice": { + "type": "object", + "required": ["id", "label"], + "properties": { + "id": { + "type": "string", + "minLength": 1, + "description": "传给上游语音接口的稳定音色 ID。" + }, + "label": { + "type": "string", + "minLength": 1, + "description": "面向用户展示的音色名称。" + } + }, + "additionalProperties": false + }, + "speechModelProfile": { + "type": "object", + "required": ["protocol", "protocolVersion", "modes"], + "properties": { + "protocol": { + "type": "string", + "minLength": 1, + "pattern": "^[a-z0-9]+(?:-[a-z0-9]+)*$", + "description": "稳定语音协议族，如 xiaomi-mimo-chat-asr；模型常规迭代不得改变此值。" + }, + "protocolVersion": { + "type": "string", + "pattern": "^[1-9][0-9]*$", + "description": "协议不兼容版本号；只有请求、认证、响应或流式状态机发生不兼容变化时才递增。" + }, + "modes": { + "type": "array", + "minItems": 1, + "uniqueItems": true, + "items": { "type": "string", "enum": ["batch", "streaming"] }, + "description": "模型协议支持的调用模式：batch（整段）和/或 streaming（流式）。" + }, + "inputFormats": { + "type": "array", + "minItems": 1, + "uniqueItems": true, + "items": { "type": "string", "minLength": 1 }, + "description": "ASR 上游接受的音频格式，如 wav、mp3、webm；格式名使用适配器可识别的规范值。" + }, + "outputFormats": { + "type": "array", + "minItems": 1, + "uniqueItems": true, + "items": { "type": "string", "minLength": 1 }, + "description": "TTS 上游可返回的音频格式，如 pcm16le、wav、mp3；运行时需与适配器和播放器能力求交集。" + }, + "sampleRates": { + "type": "array", + "minItems": 1, + "uniqueItems": true, + "items": { "type": "integer", "minimum": 1 }, + "description": "支持的采样率（Hz）；运行时不得假设固定采样率，需完成能力协商。" + }, + "defaultVoice": { + "type": "string", + "minLength": 1, + "description": "TTS 推荐默认音色 ID；声明 voices 时必须能在 voices 中找到同 ID 条目。" + }, + "voices": { + "type": "array", + "minItems": 1, + "items": { "$ref": "#/definitions/speechVoice" }, + "description": "TTS 可选音色列表，统一为稳定 id 与用户可读 label。" + }, + "languageHints": { + "type": "boolean", + "description": "ASR 协议是否接受语言提示参数；不是模型支持语言范围清单。" + } + }, + "additionalProperties": false + }, + "modelSpec": { + "type": "object", + "required": ["id", "spec"], + "properties": { + "id": { + "type": "string", + "description": "模型标准标识符，通常等于上游 model_name 原样（保留大小写与点号，如 claude-sonnet-4-6、MiniMax-M2.7）。同时作为精确匹配键；匹配时大小写/分隔符不敏感（由匹配器归一化）。", + "minLength": 1 + }, + "displayName": { + "type": "string", + "description": "模型友好显示名（如 Claude Sonnet 4.6）。命中后可覆盖 NewAPI 返回的原始 model_name，优化 UI 展示。" + }, + "family": { + "type": "string", + "description": "模型家族标识（如 claude-sonnet、gpt-5、gemini-2.5）。当精确/别名/pattern 匹配均失败时，按最长 family 前缀兜底匹配，返回保守参数并标记 inferred。" + }, + "match": { + "type": "object", + "description": "通用匹配规则。匹配器会先做归一化（小写、去 vendor 前缀如 anthropic/、去网关装饰、统一分隔符、去日期戳与波动后缀），再分层匹配：exact > 归一化 exact > 去日期戳后 exact > pattern（取最长） > family 兜底。", + "properties": { + "exact": { + "type": "array", + "description": "精确别名列表（除 id 外的已知精确名，含网关可能使用的全名）。归一化后比对。", + "items": { "type": "string", "minLength": 1 } + }, + "patterns": { + "type": "array", + "description": "glob 模式列表（如 claude-sonnet-4-6*）。仅在 exact + 去日期戳仍无法覆盖变体时使用；避免对存在子型号的父模型（如 gpt-4o 之于 gpt-4o-mini）使用宽泛通配，以防误吞。多命中取最长 pattern。", + "items": { "type": "string", "minLength": 1 } + } + }, + "additionalProperties": false + }, + "routing": { + "type": "object", + "description": "智能路由策略。存在即表示该精确模型可进入路由候选；模型能力和 reasoning 合同均以本条目为准，Provider 只能为接入面收紧。", + "required": ["tier", "routingPriority", "eligibleForAgent", "defaultReference", "reasoning"], + "properties": { + "tier": { + "type": "string", + "enum": ["flagship", "balanced", "lightweight"], + "description": "模型在智能路由中的能力量级。" + }, + "routingPriority": { + "type": "integer", + "minimum": 0, + "maximum": 10000, + "description": "同一量级内的稳定优先级，数值越小越优先。" + }, + "eligibleForAgent": { + "type": "boolean", + "description": "是否允许作为 Agent 主模型；媒体专用模型必须为 false。" + }, + "defaultReference": { + "type": "boolean", + "description": "是否是默认模型的策略参考，只影响同级排序，不是强制选择。" + }, + "reasoning": { + "type": "object", + "description": "模型级标准化 reasoning 合同。auto 为不发送参数，off 为显式关闭；Provider 接入面可收紧此集合，但不得扩张。", + "required": ["supportedModes", "defaultMode"], + "properties": { + "supportedModes": { + "type": "array", + "minItems": 1, + "uniqueItems": true, + "items": { + "type": "string", + "enum": ["auto", "off", "minimal", "low", "medium", "high", "xhigh", "max"] + }, + "description": "产品层可用的 reasoning 档位。" + }, + "defaultMode": { + "type": "string", + "enum": ["auto", "off", "minimal", "low", "medium", "high", "xhigh", "max"], + "description": "没有显式请求时的策略默认档位，必须包含在 supportedModes 中。" + } + }, + "allOf": [ + { "if": { "properties": { "defaultMode": { "const": "auto" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "auto" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "off" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "off" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "minimal" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "minimal" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "low" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "low" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "medium" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "medium" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "high" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "high" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "xhigh" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "xhigh" } } } } }, + { "if": { "properties": { "defaultMode": { "const": "max" } } }, "then": { "properties": { "supportedModes": { "contains": { "const": "max" } } } } } + ], + "additionalProperties": false + } + }, + "additionalProperties": false + }, + "spec": { + "type": "object", + "description": "模型内在参数（不含任何价格字段）", + "properties": { + "contextWindow": { + "type": "number", + "description": "上下文窗口大小（token 数）。下游用于上下文裁剪阈值计算。", + "minimum": 0 + }, + "maxOutputTokens": { + "type": "number", + "description": "单次请求最大输出 token 数。tags 字符串无法表达，是本规格库相对 NewAPI tags 的核心补充。", + "minimum": 0 + }, + "capabilities": { + "type": "array", + "description": "模型能力标签：chat、vision、tool_use、code、reasoning 等。与 NewAPI tags 解析结果取并集。", + "items": { "type": "string" } + }, + "serviceType": { + "type": ["string", "array"], + "items": { "type": "string" }, + "description": "服务类型，单个字符串或数组：chat、reasoning、fast、vision、embedding、image_gen、tts、asr 等。命中时优先于 tags 解析。" + }, + "defaultTemperature": { + "type": ["number", "null"], + "description": "推荐默认温度（0-2）。null 表示模型不支持温度调节（如 reasoning 模型），调用方应跳过该参数。本文件为新文件，老客户端不读，故允许 null。", + "minimum": 0, + "maximum": 2 + }, + "defaultTopP": { + "type": ["number", "null"], + "description": "推荐默认 Top-P（0-1）。null 表示不支持调节。", + "minimum": 0, + "maximum": 1 + }, + "supportsReasoning": { + "type": "boolean", + "description": "是否为推理模型。用于运行时决定是否启用 reasoning 相关处理。" + }, + "description": { + "type": "string", + "description": "模型简要描述（一句话特点与适用场景）。命中后可覆盖 NewAPI 原值。" + }, + "extra": { + "type": "object", + "description": "模型内在开放扩展：modelOrigin 描述真实原厂身份，speech 描述可跨小版本复用的语音协议，thinkingOnly 描述模型是否强制思考，adaptiveThinking 描述 Anthropic 模型是否只接受 adaptive thinking 线上形态；不含价格，也不得声明 Provider/接入面的 reasoning effort、thinking 回放能力（thinkingRoundTrip）或工具轮 thinking 校验口径（thinkingToolTurnValidation）——这些只能写在 provider model 的 extra 中，客户端 Schema 会拒收整份规格。", + "allOf": [ + { "not": { "required": ["reasoning"] } }, + { "not": { "required": ["thinkingRoundTrip"] } }, + { "not": { "required": ["thinkingToolTurnValidation"] } } + ], + "properties": { + "modelOrigin": { "$ref": "#/definitions/modelOrigin" }, + "speech": { "$ref": "#/definitions/speechModelProfile" }, + "samplingParametersDeprecated": { + "type": "boolean", + "description": "模型是否拒绝采样参数（如 Claude Opus 4.7+ / Opus 5 / Fable / Mythos 携带 temperature、top_p 或 top_k 即 400，Sonnet 5 对非默认值 400）。true 时客户端出站删除请求体顶层的 temperature / top_p / top_k——无论取值来自模型 defaultTemperature、用户覆写、工作流默认值还是父会话继承；false 或缺省时原样发送，客户端不按模型 ID 推断。只描述模型固有属性：精确命中的 ModelSpec 会补全到各接入面的同名模型，family 兜底不继承；作用于 Agent 引擎发出的模型请求，Claude 订阅（claude-oauth）请求由 Claude CLI 自建、不受本字段影响。" + }, + "forcedToolChoiceUnsupported": { + "type": "boolean", + "description": "模型是否拒绝强制工具选择（如 Claude Fable 5.1 / Mythos 5.1：tool_choice.type 为 any 或 tool 即 400）。true 时客户端把强制工具选择降级为 auto（Anthropic 的 any / tool、OpenAI 的 required / 指定函数），Anthropic 请求保留 disable_parallel_tool_use，none 与 auto 不受影响；false 或缺省时按调用方要求发送，客户端不按模型 ID 推断。只描述模型固有属性：精确命中的 ModelSpec 会补全到各接入面的同名模型，family 兜底不继承；作用于 Agent 引擎发出的模型请求，Claude 订阅（claude-oauth）请求由 Claude CLI 自建、不受本字段影响。" + }, + "thinkingOnly": { + "type": "boolean", + "description": "模型强制思考、无法关闭（如 GLM-5.3 系列 thinking.type 仅接受 enabled）。true 时客户端把 off 从可选档位移除，并把内部 suppressThinking 降级为 auto，避免发出必然被上游拒绝的关闭参数。只描述模型固有属性，不声明接入面支持哪些深度档位——后者只能写在 provider model 的 extra.reasoning 中。" + }, + "adaptiveThinking": { + "type": "boolean", + "description": "Anthropic Messages 模型是否属于 adaptive thinking 代际（Claude Opus 4.7+ / Opus 5 / Sonnet 5 / Fable / Mythos）。true 时客户端把思考参数发成 thinking.type=adaptive + output_config.effort，而不是会被上游 400 拒绝的 thinking.type=enabled + budget_tokens；false 表示明确不是，禁止客户端按模型 ID 推断；缺省时客户端仅在官方 api.anthropic.com 端点上按模型代际表兜底。只作用于 anthropic-messages 接入面，Claude 订阅（claude-oauth）由 Claude CLI 自行改写、不受本字段影响。命中本规格的模型由客户端 applySpec 复制到 provider model extra，运行时读的是那份副本——desirecore-cloud 等动态 Provider、非官方域名的 Anthropic Messages 中转只能从这里拿到声明，漏写即回退为 budget_tokens。match.exact、match.patterns 与 family 兜底命中时都会带上本键，因此声明本键的规格，其 exact、pattern 前缀与 family 都必须收窄到同一代际（如 family 写 claude-opus-5 而不是 claude-opus，pattern 写 claude-opus-4-7* 而不是 claude-opus-4*），否则 Opus 4.5 等旧型号会继承声明并被上游 400 拒绝。" + }, + "thinkingDefault": { + "type": "boolean", + "description": "模型默认开启思考。当前不改变客户端行为，仅供维护者参考。" + } + }, + "additionalProperties": true + }, + "releasedAt": { + "type": "string", + "description": "厂商发布年月（YYYY-MM 或 YYYY-MM-DD），辅助人工判断模型新旧。无客户端行为，仅供维护者参考何时清理该条规格。", + "pattern": "^\\d{4}-\\d{2}(-\\d{2})?$" + }, + "retiredAt": { + "type": "string", + "description": "厂商宣布停止访问的日期（YYYY-MM-DD），辅助人工批量删除已退役的规格条目。无客户端行为，仅作数据清理依据。", + "pattern": "^\\d{4}-\\d{2}-\\d{2}$" + } + }, + "additionalProperties": false + } + }, + "additionalProperties": false + } + } +} diff --git a/schemas/model-specs-index.schema.json b/schemas/model-specs-index.schema.json new file mode 100644 index 0000000..81537b1 --- /dev/null +++ b/schemas/model-specs-index.schema.json @@ -0,0 +1,88 @@ +{ + "$schema": "http://json-schema.org/draft-07/schema#", + "$id": "https://desirecore.net/schemas/config-center/model-specs-index.schema.json", + "title": "ModelSpecsIndex", + "description": "compute/model-specs 的加载顺序与智能路由三档定义。量级属于 model-specs 域，不得另建 smart-routing 目录。", + "type": "object", + "required": ["order", "routingTiers"], + "properties": { + "description": { + "type": "string", + "description": "索引说明。" + }, + "order": { + "type": "array", + "minItems": 1, + "uniqueItems": true, + "items": { "type": "string", "minLength": 1 }, + "description": "模型规格文件 basename 的稳定加载顺序。" + }, + "routingTiers": { + "type": "array", + "minItems": 3, + "maxItems": 3, + "items": { + "type": "object", + "required": ["id", "label", "labelEn", "description", "fallbackOrder"], + "properties": { + "id": { + "type": "string", + "enum": ["flagship", "balanced", "lightweight"], + "description": "量级稳定标识。" + }, + "label": { "type": "string", "minLength": 1, "description": "中文标签。" }, + "labelEn": { "type": "string", "minLength": 1, "description": "英文标签。" }, + "description": { "type": "string", "minLength": 1, "description": "适用任务说明。" }, + "fallbackOrder": { + "type": "array", + "minItems": 3, + "maxItems": 3, + "uniqueItems": true, + "items": { "type": "string", "enum": ["flagship", "balanced", "lightweight"] }, + "description": "该档位缺少可用候选时的量级降级/升级顺序，首项必须等于 id。" + } + }, + "allOf": [ + { + "if": { "properties": { "id": { "const": "flagship" } }, "required": ["id"] }, + "then": { "properties": { "fallbackOrder": { "items": [{ "const": "flagship" }] } } } + }, + { + "if": { "properties": { "id": { "const": "balanced" } }, "required": ["id"] }, + "then": { "properties": { "fallbackOrder": { "items": [{ "const": "balanced" }] } } } + }, + { + "if": { "properties": { "id": { "const": "lightweight" } }, "required": ["id"] }, + "then": { "properties": { "fallbackOrder": { "items": [{ "const": "lightweight" }] } } } + } + ], + "additionalProperties": false + }, + "description": "智能路由量级定义与回退策略。" + } + }, + "allOf": [ + { + "properties": { + "routingTiers": { + "contains": { "type": "object", "properties": { "id": { "const": "flagship" } }, "required": ["id"] } + } + } + }, + { + "properties": { + "routingTiers": { + "contains": { "type": "object", "properties": { "id": { "const": "balanced" } }, "required": ["id"] } + } + } + }, + { + "properties": { + "routingTiers": { + "contains": { "type": "object", "properties": { "id": { "const": "lightweight" } }, "required": ["id"] } + } + } + } + ], + "additionalProperties": false +}