From 73b278df51faac4737f81445368cad582ec2751d Mon Sep 17 00:00:00 2001 From: Yige Date: Mon, 5 Oct 2026 09:17:43 -0700 Subject: [PATCH] =?UTF-8?q?fix(compute):=20GLM-5.3=20=E7=B3=BB=E5=88=97?= =?UTF-8?q?=E5=9C=A8=20anthropic=20=E5=8D=8F=E8=AE=AE=E6=8C=89=E5=AE=98?= =?UTF-8?q?=E7=BD=91=E7=94=A8=20output=5Fconfig.effort=20=E6=8E=A7?= =?UTF-8?q?=E5=88=B6=E6=80=9D=E8=80=83=E5=BC=BA=E5=BA=A6=20(#120)?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit 智谱官网（国内 coding-plan/latest-model、国际 devpack/latest-model）写明 /api/anthropic 以 thinking.type 与 output_config.effort 控制思考强度，thinking.type=enabled 一律按默认 max 处理，budget_tokens 不是控制项。两个 anthropic-messages 接入面的 GLM-5.3 系列此前只有按 OpenAI 协议文档写的 low/high/max 矩阵，客户端按 budget 形态发送：low/high 在上游都成了 max， max（也是默认档）在客户端本机被拒绝；经 new-api 的 OpenAI 类渠道转换时 low 变成 reasoning_effort=medium，被智谱标准 API 拒绝。 - zhipu、zhipu-coding 接入面的 GLM-5.3 / 5.3-Flash / 5.3-FlashX 声明 adaptiveThinking - GLM-5.3 系列规格加 anthropic-messages 协议 profile（V1 副本为 spec.extra.adaptiveThinking）， 覆盖经 ModelSpec 补全的云端条目 - 来源记录补国际 Coding Plan 页；按量 API Key 与 FlashX 按同一端点推定，未列入已核字段 - presetDataVersion 130 --- __tests__/validate.test.mjs | 6 ++- catalog/v2/catalog.json | 49 +++++++++++++++++-- compute/coding-plans/zhipu-coding.json | 2 + compute/model-specs/_index.json | 2 +- compute/model-specs/zhipu.json | 3 ++ compute/providers/zhipu.json | 3 ++ compute/v2/catalog.json | 43 ++++++++++++++-- .../coding-plans/zhipu-coding.json | 2 + .../v2/compatibility/model-specs/zhipu.json | 3 ++ compute/v2/compatibility/providers/zhipu.json | 3 ++ compute/v2/manifest.json | 2 +- compute/v2/resolver-parity.json | 10 ++-- docs/model-sources/providers/zhipu/AUDIT.md | 1 + docs/model-sources/providers/zhipu/SOURCES.md | 23 +++++++-- .../providers/zhipu/models/glm-5.3-flash.md | 12 +++-- .../providers/zhipu/models/glm-5.3-flashx.md | 11 ++++- .../providers/zhipu/models/glm-5.3.md | 12 +++-- docs/source-gaps/providers/zhipu.md | 16 +++--- manifest.json | 2 +- 19 files changed, 170 insertions(+), 35 deletions(-) diff --git a/__tests__/validate.test.mjs b/__tests__/validate.test.mjs index 3acee37..9965825 100644 --- a/__tests__/validate.test.mjs +++ b/__tests__/validate.test.mjs @@ -794,7 +794,11 @@ describe('真实数据全量校验', () => { // Anthropic Messages 中转只能从 ModelSpec 补全拿到声明，预置 Provider 条目不经补全， // 两处漏写都会回退成 budget_tokens（desirecore#2940）。Opus 4.6 / Sonnet 4.6 仍接受 // budget_tokens（仅弃用），不在强制范围内。新代际若同样拒绝 budget_tokens，先扩展这里的正则。 - const adaptiveGeneration = /^claude-(?:fable|mythos)(?:$|-)|^claude-(?:opus|sonnet)-(?:[5-9]|[1-9]\d)(?:$|-)|^claude-opus-4-[78](?:$|-)|^minimax-m3-1-flash-preview$/ + // GLM-5.3 系列不拒绝 budget_tokens，但智谱官网（docs.bigmodel.cn/cn/coding-plan/latest-model、 + // docs.z.ai/devpack/latest-model）写明 /api/anthropic 以 thinking.type 与 output_config.effort 控制强度， + // thinking.type=enabled 一律按默认 max 处理，budget_tokens 不是控制项；不声明 adaptive，客户端发出的 + // 档位在上游全部变成 max，且 max 档无法发出（desirecore 档位协商回退到 max 后在本机拒绝）。 + const adaptiveGeneration = /^claude-(?:fable|mythos)(?:$|-)|^claude-(?:opus|sonnet)-(?:[5-9]|[1-9]\d)(?:$|-)|^claude-opus-4-[78](?:$|-)|^minimax-m3-1-flash-preview$|^glm-5-3(?:-flashx?)?$/ const thinkingOnlyGeneration = /^claude-(?:fable|mythos)(?:$|-)|^claude-opus-5-5(?:$|-)/ // 与客户端匹配器同口径：小写、去 vendor 前缀、统一分隔符 const normalizeId = (id) => id.toLowerCase().trim().split('/').at(-1).replace(/[._:\s]+/g, '-') diff --git a/catalog/v2/catalog.json b/catalog/v2/catalog.json index 03aa31f..06e7ad3 100644 --- a/catalog/v2/catalog.json +++ b/catalog/v2/catalog.json @@ -14676,6 +14676,7 @@ "extra" ], "extraKeys": [ + "adaptiveThinking", "thinkingOnly" ] }, @@ -14690,7 +14691,17 @@ ] } } - } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] }, { "modelRef": "zhipu:glm-5.3", @@ -14759,6 +14770,7 @@ "releasedAt" ], "extraKeys": [ + "adaptiveThinking", "thinkingDefault", "thinkingOnly" ] @@ -14777,7 +14789,17 @@ "recordedFacts": { "releasedAt": "2026-08-16" } - } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] }, { "modelRef": "zhipu:glm-5.3-flash", @@ -14849,6 +14871,7 @@ "releasedAt" ], "extraKeys": [ + "adaptiveThinking", "thinkingDefault", "thinkingOnly" ] @@ -14867,7 +14890,17 @@ "recordedFacts": { "releasedAt": "2026-08-26" } - } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] }, { "modelRef": "zhipu:glm-5.2", @@ -34946,6 +34979,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -35013,6 +35047,7 @@ "outputPrice" ], "extraKeys": [ + "adaptiveThinking", "thinkingOnly", "cacheHitPrice", "reasoning", @@ -35031,6 +35066,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -35102,6 +35138,7 @@ "outputPrice" ], "extraKeys": [ + "adaptiveThinking", "thinkingDefault", "thinkingOnly", "cacheHitPrice", @@ -35121,6 +35158,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -35192,6 +35230,7 @@ "outputPrice" ], "extraKeys": [ + "adaptiveThinking", "thinkingDefault", "thinkingOnly", "cacheHitPrice", @@ -41080,6 +41119,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -41128,6 +41168,7 @@ "source" ], "extraKeys": [ + "adaptiveThinking", "thinkingDefault", "thinkingOnly", "reasoning", @@ -41166,6 +41207,7 @@ } }, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -41207,6 +41249,7 @@ "source" ], "extraKeys": [ + "adaptiveThinking", "supportsThinking", "thinkingDefault", "thinkingOnly", diff --git a/compute/coding-plans/zhipu-coding.json b/compute/coding-plans/zhipu-coding.json index 0b4951d..9b084e8 100644 --- a/compute/coding-plans/zhipu-coding.json +++ b/compute/coding-plans/zhipu-coding.json @@ -50,6 +50,7 @@ "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": { @@ -92,6 +93,7 @@ "defaultTemperature": 1, "defaultTopP": 0.95, "extra": { + "adaptiveThinking": true, "supportsThinking": true, "thinkingDefault": true, "thinkingOnly": true, diff --git a/compute/model-specs/_index.json b/compute/model-specs/_index.json index b53d351..9fe2b20 100644 --- a/compute/model-specs/_index.json +++ b/compute/model-specs/_index.json @@ -1,5 +1,5 @@ { - "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 129：按官网补对话规格的 tool_use 标签，路由策略不变；此前批次 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。", + "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", diff --git a/compute/model-specs/zhipu.json b/compute/model-specs/zhipu.json index 94d0d74..4351bd9 100644 --- a/compute/model-specs/zhipu.json +++ b/compute/model-specs/zhipu.json @@ -33,6 +33,7 @@ "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 } }, @@ -84,6 +85,7 @@ "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 }, @@ -140,6 +142,7 @@ "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 }, diff --git a/compute/providers/zhipu.json b/compute/providers/zhipu.json index bf391da..20245c4 100644 --- a/compute/providers/zhipu.json +++ b/compute/providers/zhipu.json @@ -48,6 +48,7 @@ ], "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": { @@ -89,6 +90,7 @@ "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, @@ -128,6 +130,7 @@ "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, diff --git a/compute/v2/catalog.json b/compute/v2/catalog.json index 32ef589..49904dd 100644 --- a/compute/v2/catalog.json +++ b/compute/v2/catalog.json @@ -9342,7 +9342,17 @@ "metadata": { "recordedExtra": {}, "sourceFile": "compute/model-specs/zhipu.json" - } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] }, { "modelRef": "zhipu:glm-5.3", @@ -9401,7 +9411,17 @@ "recordedFacts": { "releasedAt": "2026-08-16" } - } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] }, { "modelRef": "zhipu:glm-5.3-flash", @@ -9463,7 +9483,17 @@ "recordedFacts": { "releasedAt": "2026-08-26" } - } + }, + "protocolProfiles": [ + { + "apiFormats": [ + "anthropic-messages" + ], + "request": { + "adaptiveThinking": true + } + } + ] }, { "modelRef": "zhipu:glm-5.2", @@ -25139,6 +25169,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -25204,6 +25235,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -25271,6 +25303,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -29991,6 +30024,7 @@ }, "overrides": {}, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -30056,6 +30090,7 @@ } }, "request": { + "adaptiveThinking": true, "constraints": { "thinkingOnly": true }, @@ -30304,5 +30339,5 @@ "authority": "catalog/v2/catalog.json", "verification": "migration-preserves-existing-evidence" }, - "digest": "7ed3f47d0c41bec661509d5cb76d8130dbd4fb1e63d14eae0e73de5f00bcc4cb" + "digest": "8f0c6ee356bb44e60fb17d1dca04b012d47cae9cea9fd7890f874894ee74cdb5" } diff --git a/compute/v2/compatibility/coding-plans/zhipu-coding.json b/compute/v2/compatibility/coding-plans/zhipu-coding.json index 0b4951d..9b084e8 100644 --- a/compute/v2/compatibility/coding-plans/zhipu-coding.json +++ b/compute/v2/compatibility/coding-plans/zhipu-coding.json @@ -50,6 +50,7 @@ "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": { @@ -92,6 +93,7 @@ "defaultTemperature": 1, "defaultTopP": 0.95, "extra": { + "adaptiveThinking": true, "supportsThinking": true, "thinkingDefault": true, "thinkingOnly": true, diff --git a/compute/v2/compatibility/model-specs/zhipu.json b/compute/v2/compatibility/model-specs/zhipu.json index bd96cd1..5f44602 100644 --- a/compute/v2/compatibility/model-specs/zhipu.json +++ b/compute/v2/compatibility/model-specs/zhipu.json @@ -33,6 +33,7 @@ "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 } }, @@ -84,6 +85,7 @@ "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 }, @@ -140,6 +142,7 @@ "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 }, diff --git a/compute/v2/compatibility/providers/zhipu.json b/compute/v2/compatibility/providers/zhipu.json index bf391da..20245c4 100644 --- a/compute/v2/compatibility/providers/zhipu.json +++ b/compute/v2/compatibility/providers/zhipu.json @@ -48,6 +48,7 @@ ], "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": { @@ -89,6 +90,7 @@ "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, @@ -128,6 +130,7 @@ "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, diff --git a/compute/v2/manifest.json b/compute/v2/manifest.json index ac74dfb..1fe31cc 100644 --- a/compute/v2/manifest.json +++ b/compute/v2/manifest.json @@ -3,5 +3,5 @@ "state": "candidate", "requiredClientVersion": null, "catalogPath": "compute/v2/catalog.json", - "digest": "7ed3f47d0c41bec661509d5cb76d8130dbd4fb1e63d14eae0e73de5f00bcc4cb" + "digest": "8f0c6ee356bb44e60fb17d1dca04b012d47cae9cea9fd7890f874894ee74cdb5" } diff --git a/compute/v2/resolver-parity.json b/compute/v2/resolver-parity.json index 3ed5034..efdf23b 100644 --- a/compute/v2/resolver-parity.json +++ b/compute/v2/resolver-parity.json @@ -2974,7 +2974,7 @@ { "accessId": "provider-zhipu-001", "apiModelId": "glm-5.3-flashx", - "digest": "ea21bf8d96f65888a4f6991ed37a7824894d94fd1baa8509ccd9a4c9cea5c51c", + "digest": "d4886f90bf9c291fe3c23183348e995ad9a09d469b235ef0dc662e133a62d350", "selectable": true, "effectiveApiFormat": "anthropic-messages", "effectiveApiModelId": "glm-5.3-flashx", @@ -2986,7 +2986,7 @@ { "accessId": "provider-zhipu-001", "apiModelId": "glm-5.3-flash", - "digest": "0e50b1f0393cb2fd782ffa161b65c6151e213dcd7f9f4eae8621e4fdb1f4915d", + "digest": "a06221f39d4d13993cf160510bfe6dc2be5c28d082f6df7bd2ff81ce403d5b63", "selectable": true, "effectiveApiFormat": "anthropic-messages", "effectiveApiModelId": "glm-5.3-flash", @@ -2998,7 +2998,7 @@ { "accessId": "provider-zhipu-001", "apiModelId": "glm-5.3", - "digest": "44f0f52f12351b78825f7e32930e8e9f81693a80a3a277c196c30f88498e500d", + "digest": "d9c73c7eac661e0f6a429ead1ea86fb8920875b71c9577ec01c2f9f253cecb7b", "selectable": true, "effectiveApiFormat": "anthropic-messages", "effectiveApiModelId": "glm-5.3", @@ -4100,7 +4100,7 @@ { "accessId": "provider-zhipu-coding-001", "apiModelId": "glm-5.3", - "digest": "215a2293b36c7ba03f525fa42914eb81425051375a8b2f629bc76affc3d04691", + "digest": "ac57a48657071e66a8d0253e284b5cc598ae9bcf85d63ecde8e9d3672d17e8f8", "selectable": true, "effectiveApiFormat": "anthropic-messages", "effectiveApiModelId": "glm-5.3", @@ -4112,7 +4112,7 @@ { "accessId": "provider-zhipu-coding-001", "apiModelId": "glm-5.3-flash", - "digest": "c254302f180a5afb2303be38a65eb61237b464a4e4456b56297dcc9c2d2da38f", + "digest": "b44eecd20c164d0e36e7a54700f9f9505deb5d435b54532c25af9d3d899fdfa3", "selectable": true, "effectiveApiFormat": "anthropic-messages", "effectiveApiModelId": "glm-5.3-flash", diff --git a/docs/model-sources/providers/zhipu/AUDIT.md b/docs/model-sources/providers/zhipu/AUDIT.md index 74152da..978b285 100644 --- a/docs/model-sources/providers/zhipu/AUDIT.md +++ b/docs/model-sources/providers/zhipu/AUDIT.md @@ -10,6 +10,7 @@ | GLM5.3/Flash精确窗口 | 官网1M/128K简写尚不足以证明本仓具体整数，取得明文限制后统一 | | 存量Embedding/多模态接口 | 按独立接口读取向量维度、服务单位、端点和兼容格式；不由Chat参数迁移 | | 国内价与国际价 | 分币种/地域与峰谷/缓存，不能相互覆盖；逐型号补字段来源 | +| 按量 API Key 与 FlashX 的 Anthropic 思考强度 | 2026-10-05 已按 Coding Plan 官网页给 GLM-5.3 系列声明 `adaptiveThinking`（客户端改发 `output_config.effort`）；官网只在 Coding Plan 页说明 /api/anthropic 的处理。需用按量 API Key 对 GLM-5.3 与 FlashX 各发一次 `output_config.effort` 为 low 与 max 的请求，确认均被接受且思考长度有差异，再把 `zhipu` 接入面与 FlashX 规格的 `adaptiveThinking` 列入已核字段 | 未进行账号调用，不宣称国际套餐已实测。 diff --git a/docs/model-sources/providers/zhipu/SOURCES.md b/docs/model-sources/providers/zhipu/SOURCES.md index 5be64e5..8668efd 100644 --- a/docs/model-sources/providers/zhipu/SOURCES.md +++ b/docs/model-sources/providers/zhipu/SOURCES.md @@ -21,7 +21,12 @@ ### glm-plan - 入口：[glm-plan](https://docs.bigmodel.cn/cn/coding-plan/latest-model.md) -- 类型：`official-doc`；当前读取状态：`fetched`；地域/接入面：`中国区 GLM Coding Plan 文档；国际 api.z.ai 可用性另核`。 +- 类型：`official-doc`；当前读取状态：`fetched`；地域/接入面：`中国区 GLM Coding Plan 文档（Claude Code 走 open.bigmodel.cn/api/anthropic）；国际 api.z.ai 可用性另核`。 + +### glm-plan-intl + +- 入口：[glm-plan-intl](https://docs.z.ai/devpack/latest-model.md) +- 类型：`official-doc`；当前读取状态：`fetched`；地域/接入面：`Z.AI 国际 GLM Coding Plan 文档（Claude Code 走 api.z.ai/api/anthropic）`。 ### glm52-cn-api @@ -52,6 +57,8 @@ 国内官方价格页确认 5.3/Flash/FlashX 为 8/28、0.8/2.8、2/7 元/百万 tokens。官网给出的窗口简写 1M/128K 未直接作为跨接入面字节级精确值证明；本次不改已有换算。低/高/max、默认 max 和始终思考已核对。Coding Plan 提供 5.3 与 Flash，不提供 FlashX；国际套餐端点的账号实际可用性仍需验证。 +2026-10-05 Anthropic 协议思考强度核对：`glm-plan`（国内）与 `glm-plan-intl`（国际）两页「如何切换 effort」一节明文写出：Claude Code 使用 `thinking.type` 与 `output_config.effort`，Codex 使用 `reasoning.effort`；优先级为显式 effort > thinking 开关 > 默认 max；`thinking.type` 未传 / `true` / `enabled` / `adaptive` 按默认 max 处理，`false` / `disabled` / `none` / `off` 转为 low；effort `minimal` / `light` / `low` → low，`medium` / `high` → high，`xhigh` / `max` / `ultra` → max，其他未知值回退 max。`budget_tokens` 不在这张表的控制项里。据此在两个 anthropic-messages 接入面（`zhipu`、`zhipu-coding`）的 GLM-5.3 系列条目与 GLM-5.3 系列规格的 anthropic-messages 协议 profile 上声明 `adaptiveThinking`，客户端改用 `output_config.effort` 表达档位。两页属于 Coding Plan 文档：`zhipu-coding` 的 GLM-5.3 / 5.3-Flash 与规格中的这两个型号直接对应；`zhipu`（按量 API Key）与 FlashX 按同一 /api/anthropic 端点推定，API Key 账号实测待补，故未列入这些条目的已核字段。官网 Claude API 兼容页（`/cn/guide/develop/claude/introduction`）没有列出思考参数。 + 2026-10-05 工具调用核对：以上 `zhipu-glm-4-6-page`、`zhipu-glm-4-7-page`、`zhipu-model-overview` 只用于核对函数 / 工具调用是否受支持，读取日期 2026-10-05；没有借此复核窗口、价格等其他字段，也没有做账号调用验收。能力标签里只有 `tool_use` 一项以这些来源为依据。 状态 `partial` 仅表示列出的字段已核实；不能推断整条配置、价格、账号权限或真实模型调用都已验收。`pending` 没有已核字段；`historical` 只作历史兼容参考。路由 tier/priority、产品标签和默认选择属于本仓策略，不伪装成官网数据。 @@ -97,12 +104,22 @@ "id": "glm-plan", "url": "https://docs.bigmodel.cn/cn/coding-plan/latest-model.md", "kind": "official-doc", - "scope": "中国区 GLM Coding Plan 文档；国际 api.z.ai 可用性另核", + "scope": "中国区 GLM Coding Plan 文档（Claude Code 走 open.bigmodel.cn/api/anthropic）；国际 api.z.ai 可用性另核", "retrieval": "fetched", - "checkedAt": "2026-10-03", + "checkedAt": "2026-10-05", "contentSha256": "d5c5eb7ef0ddf5a89128018fffb1d96a7e4a6745c8c16dc53c5c4453e59891fc", "resolvedUrl": "https://docs.bigmodel.cn/cn/coding-plan/latest-model.md" }, + { + "id": "glm-plan-intl", + "url": "https://docs.z.ai/devpack/latest-model.md", + "kind": "official-doc", + "scope": "Z.AI 国际 GLM Coding Plan 文档（Claude Code 走 api.z.ai/api/anthropic）", + "retrieval": "fetched", + "checkedAt": "2026-10-05", + "contentSha256": "4cc2c1a526a96efbe4c9d90be887fd07dc93a49d6c23b2cd1884899f822aea43", + "resolvedUrl": "https://docs.z.ai/devpack/latest-model.md" + }, { "id": "glm52-cn-api", "url": "https://docs.bigmodel.cn/api-reference/%E6%A8%A1%E5%9E%8B-api/%E5%AF%B9%E8%AF%9D%E8%A1%A5%E5%85%A8.md", diff --git a/docs/model-sources/providers/zhipu/models/glm-5.3-flash.md b/docs/model-sources/providers/zhipu/models/glm-5.3-flash.md index 303fd6c..97ff99c 100644 --- a/docs/model-sources/providers/zhipu/models/glm-5.3-flash.md +++ b/docs/model-sources/providers/zhipu/models/glm-5.3-flash.md @@ -4,6 +4,7 @@ - 精确模型 ID：`glm-5.3-flash`。 - 核验日期：2026-10-03；本次只迁移记录，没有重新核验或提高状态。 +- 2026-10-05 补核：Anthropic 协议的思考强度控制（`thinking.type` 与 `output_config.effort`），见官网证据目录「已知边界与核验说明」；其余字段未重新核验。 ## 适用接入面 @@ -55,6 +56,7 @@ ], "defaultEffort": "max" }, + "extra.adaptiveThinking": true, "extra.thinkingOnly": true, "extra.cacheHitPrice": 0.23, "extra.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。" @@ -62,11 +64,13 @@ ``` +`extra.adaptiveThinking` 依据同一 `/api/anthropic` 端点在 Coding Plan 官网页（[glm-plan](../SOURCES.md#glm-plan)）上的思考强度说明推定；按量 API Key 的账号实测待补，故未列入本接入面已核字段。 + | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3-flash` | 1048576 / 131072 | CNY：0.8 / 2.8 | `extra.cacheHitPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `extra.reasoning.defaultEffort`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.reasoning.supportedEfforts`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash), `inputPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `modelName`→[glm53-flash](../SOURCES.md#glm53-flash), `outputPrice`→[glm-pricing](../SOURCES.md#glm-pricing) | [glm-pricing](../SOURCES.md#glm-pricing), [glm53-flash](../SOURCES.md#glm53-flash) | partial | `7ac4fc2d14893734f4e120e594d3f7a434166e27e64fc2f34da94fbb6296c9fe` | +| `glm-5.3-flash` | 1048576 / 131072 | CNY：0.8 / 2.8 | `extra.cacheHitPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `extra.reasoning.defaultEffort`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.reasoning.supportedEfforts`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash), `inputPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `modelName`→[glm53-flash](../SOURCES.md#glm53-flash), `outputPrice`→[glm-pricing](../SOURCES.md#glm-pricing) | [glm-pricing](../SOURCES.md#glm-pricing), [glm53-flash](../SOURCES.md#glm53-flash) | partial | `e9b5a587cd462e819120f6771f9068aecc4efb5d6381823ecfffbd2a7a4a8e22` | ### compute/coding-plans/zhipu-coding.json @@ -108,6 +112,7 @@ ], "defaultEffort": "max" }, + "extra.adaptiveThinking": true, "extra.thinkingOnly": true } ``` @@ -117,7 +122,7 @@ | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3-flash` | 1048576 / 131072 | 套餐：未声明 / 未声明 | `extra.reasoning.defaultEffort`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.reasoning.supportedEfforts`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash), `modelName`→[glm53-flash](../SOURCES.md#glm53-flash) | [glm53-flash](../SOURCES.md#glm53-flash) | partial | `38f4c09e9f3aa0fd17481b40d42378de63362ebbf714ae82222dbcbd0510e7f4` | +| `glm-5.3-flash` | 1048576 / 131072 | 套餐：未声明 / 未声明 | `extra.adaptiveThinking`→[glm-plan-intl](../SOURCES.md#glm-plan-intl), `extra.reasoning.defaultEffort`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.reasoning.supportedEfforts`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash), `modelName`→[glm53-flash](../SOURCES.md#glm53-flash) | [glm-plan-intl](../SOURCES.md#glm-plan-intl), [glm53-flash](../SOURCES.md#glm53-flash) | partial | `72a649e6a18df9559bb6606103c29baf303e045d11037c1765f116d50f04d223` | ### compute/model-specs/zhipu.json @@ -151,6 +156,7 @@ "spec.defaultTemperature": 1, "spec.defaultTopP": 0.95, "spec.releasedAt": "2026-08-26", + "spec.extra.adaptiveThinking": true, "spec.extra.thinkingOnly": true, "routing.tier": "balanced", "routing.routingPriority": 48, @@ -172,7 +178,7 @@ | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3-flash` | 1048576 / 131072 | 非计价主数据 | `id`→[glm53-flash](../SOURCES.md#glm53-flash), `spec.extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash) | [glm53-flash](../SOURCES.md#glm53-flash) | partial | `28114bb8b0a40a61668f84c0edbefc7a70d5dba44102ae184b16098798fa0609` | +| `glm-5.3-flash` | 1048576 / 131072 | 非计价主数据 | `id`→[glm53-flash](../SOURCES.md#glm53-flash), `spec.extra.adaptiveThinking`→[glm-plan](../SOURCES.md#glm-plan), `spec.extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash) | [glm-plan](../SOURCES.md#glm-plan), [glm53-flash](../SOURCES.md#glm53-flash) | partial | `a9447c04b737184ab52b7aaac1380580422f05ade4cfb982675d02d6c31d6894` | ## 下次更新核查 diff --git a/docs/model-sources/providers/zhipu/models/glm-5.3-flashx.md b/docs/model-sources/providers/zhipu/models/glm-5.3-flashx.md index 4fb933d..a571aa1 100644 --- a/docs/model-sources/providers/zhipu/models/glm-5.3-flashx.md +++ b/docs/model-sources/providers/zhipu/models/glm-5.3-flashx.md @@ -4,6 +4,7 @@ - 精确模型 ID：`glm-5.3-flashx`。 - 核验日期：2026-10-03；本次只迁移记录，没有重新核验或提高状态。 +- 2026-10-05 补核：Anthropic 协议的思考强度控制（`thinking.type` 与 `output_config.effort`），见官网证据目录「已知边界与核验说明」；其余字段未重新核验。 ## 适用接入面 @@ -51,6 +52,7 @@ ], "defaultEffort": "max" }, + "extra.adaptiveThinking": true, "extra.thinkingOnly": true, "extra.cacheHitPrice": 0.57, "extra.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。" @@ -58,11 +60,13 @@ ``` +`extra.adaptiveThinking` 依据同一 `/api/anthropic` 端点在 Coding Plan 官网页（[glm-plan](../SOURCES.md#glm-plan)）上的思考强度说明推定；按量 API Key 的账号实测待补，故未列入本接入面已核字段。 + | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3-flashx` | 1000000 / 128000 | CNY：2 / 7 | `extra.cacheHitPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `extra.reasoning.defaultEffort`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.reasoning.supportedEfforts`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash), `inputPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `modelName`→[glm53-flash](../SOURCES.md#glm53-flash), `outputPrice`→[glm-pricing](../SOURCES.md#glm-pricing) | [glm-pricing](../SOURCES.md#glm-pricing), [glm53-flash](../SOURCES.md#glm53-flash) | partial | `90198f90256cc40ac50dd2ac6f395ba1a0f6062cdefd392f61840c1569b6f8e1` | +| `glm-5.3-flashx` | 1000000 / 128000 | CNY：2 / 7 | `extra.cacheHitPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `extra.reasoning.defaultEffort`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.reasoning.supportedEfforts`→[glm53-flash](../SOURCES.md#glm53-flash), `extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash), `inputPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `modelName`→[glm53-flash](../SOURCES.md#glm53-flash), `outputPrice`→[glm-pricing](../SOURCES.md#glm-pricing) | [glm-pricing](../SOURCES.md#glm-pricing), [glm53-flash](../SOURCES.md#glm53-flash) | partial | `22d007a5ca6b209ac640ba425414039efb13bbae89294c0722e6d571933c47ba` | ### compute/model-specs/zhipu.json @@ -92,6 +96,7 @@ "fast" ], "spec.supportsReasoning": true, + "spec.extra.adaptiveThinking": true, "spec.extra.thinkingOnly": true, "routing.tier": "balanced", "routing.routingPriority": 49, @@ -109,11 +114,13 @@ ``` +`spec.extra.adaptiveThinking` 是 anthropic-messages 协议 profile 的 V1 副本；Coding Plan 官网页未覆盖 FlashX，按同一端点推定，未列入已核字段。 + | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3-flashx` | 1000000 / 128000 | 非计价主数据 | `id`→[glm53-flash](../SOURCES.md#glm53-flash), `spec.extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash) | [glm53-flash](../SOURCES.md#glm53-flash) | partial | `6d10e6cf17f4461142a1827d036c3cb2b7d798d5e579398e1d4ca9bc4d00394b` | +| `glm-5.3-flashx` | 1000000 / 128000 | 非计价主数据 | `id`→[glm53-flash](../SOURCES.md#glm53-flash), `spec.extra.thinkingOnly`→[glm53-flash](../SOURCES.md#glm53-flash) | [glm53-flash](../SOURCES.md#glm53-flash) | partial | `29f64f9aa2a973bc71f00b2c053509b23ae8af5654142337c0bc99fc05b0ba89` | ## 下次更新核查 diff --git a/docs/model-sources/providers/zhipu/models/glm-5.3.md b/docs/model-sources/providers/zhipu/models/glm-5.3.md index f025288..7030217 100644 --- a/docs/model-sources/providers/zhipu/models/glm-5.3.md +++ b/docs/model-sources/providers/zhipu/models/glm-5.3.md @@ -4,6 +4,7 @@ - 精确模型 ID：`glm-5.3`。 - 核验日期：2026-10-03；本次只迁移记录，没有重新核验或提高状态。 +- 2026-10-05 补核：Anthropic 协议的思考强度控制（`thinking.type` 与 `output_config.effort`），见官网证据目录「已知边界与核验说明」；其余字段未重新核验。 ## 适用接入面 @@ -52,6 +53,7 @@ ], "defaultEffort": "max" }, + "extra.adaptiveThinking": true, "extra.thinkingOnly": true, "extra.cacheHitPrice": 2, "extra.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。" @@ -59,11 +61,13 @@ ``` +`extra.adaptiveThinking` 依据同一 `/api/anthropic` 端点在 Coding Plan 官网页（[glm-plan](../SOURCES.md#glm-plan)）上的思考强度说明推定；按量 API Key 的账号实测待补，故未列入本接入面已核字段。 + | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3` | 1048576 / 131072 | CNY：8 / 28 | `extra.cacheHitPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `extra.reasoning.defaultEffort`→[glm53](../SOURCES.md#glm53), `extra.reasoning.supportedEfforts`→[glm53](../SOURCES.md#glm53), `extra.thinkingOnly`→[glm53](../SOURCES.md#glm53), `inputPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `modelName`→[glm53](../SOURCES.md#glm53), `outputPrice`→[glm-pricing](../SOURCES.md#glm-pricing) | [glm-pricing](../SOURCES.md#glm-pricing), [glm53](../SOURCES.md#glm53) | partial | `4db175a86ebfcecbf4dacc3c3f5efe58b10d136aa09d8cac190c955b47fc550b` | +| `glm-5.3` | 1048576 / 131072 | CNY：8 / 28 | `extra.cacheHitPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `extra.reasoning.defaultEffort`→[glm53](../SOURCES.md#glm53), `extra.reasoning.supportedEfforts`→[glm53](../SOURCES.md#glm53), `extra.thinkingOnly`→[glm53](../SOURCES.md#glm53), `inputPrice`→[glm-pricing](../SOURCES.md#glm-pricing), `modelName`→[glm53](../SOURCES.md#glm53), `outputPrice`→[glm-pricing](../SOURCES.md#glm-pricing) | [glm-pricing](../SOURCES.md#glm-pricing), [glm53](../SOURCES.md#glm53) | partial | `cf4e9caea8bbef44a786a83758824551a9f4bfb0276d7ad11ce777b0aa315471` | ### compute/coding-plans/zhipu-coding.json @@ -100,6 +104,7 @@ ], "defaultEffort": "max" }, + "extra.adaptiveThinking": true, "extra.thinkingOnly": true, "extra.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。" } @@ -110,7 +115,7 @@ | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3` | 1048576 / 131072 | 套餐：未声明 / 未声明 | `extra.reasoning.defaultEffort`→[glm53](../SOURCES.md#glm53), `extra.reasoning.supportedEfforts`→[glm53](../SOURCES.md#glm53), `extra.thinkingOnly`→[glm53](../SOURCES.md#glm53), `modelName`→[glm53](../SOURCES.md#glm53) | [glm53](../SOURCES.md#glm53) | partial | `639b9e9b3d7f8eff6d533c1b36f09ea7bea64f09bc4f20157fd01e779cda1016` | +| `glm-5.3` | 1048576 / 131072 | 套餐：未声明 / 未声明 | `extra.adaptiveThinking`→[glm-plan-intl](../SOURCES.md#glm-plan-intl), `extra.reasoning.defaultEffort`→[glm53](../SOURCES.md#glm53), `extra.reasoning.supportedEfforts`→[glm53](../SOURCES.md#glm53), `extra.thinkingOnly`→[glm53](../SOURCES.md#glm53), `modelName`→[glm53](../SOURCES.md#glm53) | [glm-plan-intl](../SOURCES.md#glm-plan-intl), [glm53](../SOURCES.md#glm53) | partial | `68c44eb4198ed090d290f757b01cd7bbb95454061a04283a81eaf83f47845faa` | ### compute/model-specs/zhipu.json @@ -141,6 +146,7 @@ "spec.defaultTemperature": 1, "spec.defaultTopP": 0.95, "spec.releasedAt": "2026-08-16", + "spec.extra.adaptiveThinking": true, "spec.extra.thinkingOnly": true, "routing.tier": "flagship", "routing.routingPriority": 50, @@ -162,7 +168,7 @@ | 模型 ID | 上下文 / 输出 | 计价 | 已核字段→来源 | 来源入口 | 状态 | 数据指纹 | | --- | --- | --- | --- | --- | --- | --- | -| `glm-5.3` | 1048576 / 131072 | 非计价主数据 | `id`→[glm53](../SOURCES.md#glm53), `spec.extra.thinkingOnly`→[glm53](../SOURCES.md#glm53) | [glm53](../SOURCES.md#glm53) | partial | `a626e0690b06921d5f41fcf764a7e295e75acface2d3ed005f664aaacc38411d` | +| `glm-5.3` | 1048576 / 131072 | 非计价主数据 | `id`→[glm53](../SOURCES.md#glm53), `spec.extra.adaptiveThinking`→[glm-plan](../SOURCES.md#glm-plan), `spec.extra.thinkingOnly`→[glm53](../SOURCES.md#glm53) | [glm-plan](../SOURCES.md#glm-plan), [glm53](../SOURCES.md#glm53) | partial | `961ed39ff97065d0420ca60fc3ecf25ee605bde65939bd740362fc39726c5a5b` | ## 下次更新核查 diff --git a/docs/source-gaps/providers/zhipu.md b/docs/source-gaps/providers/zhipu.md index 25ffd07..bda9f83 100644 --- a/docs/source-gaps/providers/zhipu.md +++ b/docs/source-gaps/providers/zhipu.md @@ -49,14 +49,14 @@ | --- | --- | --- | --- | | [`glm-5.2`](../../model-sources/providers/zhipu/models/glm-5.2.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`extra.cacheHitPrice`、`extra.cacheStoragePrice`、`inputPrice`、`maxOutputTokens`、`modelName`、`outputPrice` | [glm52-cn-api](../../model-sources/providers/zhipu/SOURCES.md#glm52-cn-api)：fetched | | [`glm-5.2`](../../model-sources/providers/zhipu/models/glm-5.2.md) | [`compute/coding-plans/zhipu-coding.json`](../../../compute/coding-plans/zhipu-coding.json) | `contextWindow`、`maxOutputTokens`、`modelName` | [glm52-intl-api](../../model-sources/providers/zhipu/SOURCES.md#glm52-intl-api)：fetched | -| [`glm-5.3-flash`](../../model-sources/providers/zhipu/models/glm-5.3-flash.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`maxOutputTokens` | [glm-pricing](../../model-sources/providers/zhipu/SOURCES.md#glm-pricing)：fetched；[glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | -| [`glm-5.3-flash`](../../model-sources/providers/zhipu/models/glm-5.3-flash.md) | [`compute/coding-plans/zhipu-coding.json`](../../../compute/coding-plans/zhipu-coding.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`maxOutputTokens` | [glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | -| [`glm-5.3-flash`](../../model-sources/providers/zhipu/models/glm-5.3-flash.md) | [`compute/model-specs/zhipu.json`](../../../compute/model-specs/zhipu.json) | `spec.contextWindow`、`spec.defaultTemperature`、`spec.defaultTopP`、`spec.maxOutputTokens`、`spec.supportsReasoning` | [glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | -| [`glm-5.3-flashx`](../../model-sources/providers/zhipu/models/glm-5.3-flashx.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`maxOutputTokens` | [glm-pricing](../../model-sources/providers/zhipu/SOURCES.md#glm-pricing)：fetched；[glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | -| [`glm-5.3-flashx`](../../model-sources/providers/zhipu/models/glm-5.3-flashx.md) | [`compute/model-specs/zhipu.json`](../../../compute/model-specs/zhipu.json) | `spec.contextWindow`、`spec.maxOutputTokens`、`spec.supportsReasoning` | [glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | -| [`glm-5.3`](../../model-sources/providers/zhipu/models/glm-5.3.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`maxOutputTokens` | [glm-pricing](../../model-sources/providers/zhipu/SOURCES.md#glm-pricing)：fetched；[glm53](../../model-sources/providers/zhipu/SOURCES.md#glm53)：fetched | -| [`glm-5.3`](../../model-sources/providers/zhipu/models/glm-5.3.md) | [`compute/coding-plans/zhipu-coding.json`](../../../compute/coding-plans/zhipu-coding.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`maxOutputTokens` | [glm53](../../model-sources/providers/zhipu/SOURCES.md#glm53)：fetched | -| [`glm-5.3`](../../model-sources/providers/zhipu/models/glm-5.3.md) | [`compute/model-specs/zhipu.json`](../../../compute/model-specs/zhipu.json) | `spec.contextWindow`、`spec.defaultTemperature`、`spec.defaultTopP`、`spec.maxOutputTokens`、`spec.supportsReasoning` | [glm53](../../model-sources/providers/zhipu/SOURCES.md#glm53)：fetched | +| [`glm-5.3-flash`](../../model-sources/providers/zhipu/models/glm-5.3-flash.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`extra.adaptiveThinking`、`maxOutputTokens` | [glm-pricing](../../model-sources/providers/zhipu/SOURCES.md#glm-pricing)：fetched；[glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | +| [`glm-5.3-flash`](../../model-sources/providers/zhipu/models/glm-5.3-flash.md) | [`compute/coding-plans/zhipu-coding.json`](../../../compute/coding-plans/zhipu-coding.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`maxOutputTokens` | [glm-plan-intl](../../model-sources/providers/zhipu/SOURCES.md#glm-plan-intl)：fetched；[glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | +| [`glm-5.3-flash`](../../model-sources/providers/zhipu/models/glm-5.3-flash.md) | [`compute/model-specs/zhipu.json`](../../../compute/model-specs/zhipu.json) | `spec.contextWindow`、`spec.defaultTemperature`、`spec.defaultTopP`、`spec.maxOutputTokens`、`spec.supportsReasoning` | [glm-plan](../../model-sources/providers/zhipu/SOURCES.md#glm-plan)：fetched；[glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | +| [`glm-5.3-flashx`](../../model-sources/providers/zhipu/models/glm-5.3-flashx.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`extra.adaptiveThinking`、`maxOutputTokens` | [glm-pricing](../../model-sources/providers/zhipu/SOURCES.md#glm-pricing)：fetched；[glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | +| [`glm-5.3-flashx`](../../model-sources/providers/zhipu/models/glm-5.3-flashx.md) | [`compute/model-specs/zhipu.json`](../../../compute/model-specs/zhipu.json) | `spec.contextWindow`、`spec.extra.adaptiveThinking`、`spec.maxOutputTokens`、`spec.supportsReasoning` | [glm53-flash](../../model-sources/providers/zhipu/SOURCES.md#glm53-flash)：fetched | +| [`glm-5.3`](../../model-sources/providers/zhipu/models/glm-5.3.md) | [`compute/providers/zhipu.json`](../../../compute/providers/zhipu.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`extra.adaptiveThinking`、`maxOutputTokens` | [glm-pricing](../../model-sources/providers/zhipu/SOURCES.md#glm-pricing)：fetched；[glm53](../../model-sources/providers/zhipu/SOURCES.md#glm53)：fetched | +| [`glm-5.3`](../../model-sources/providers/zhipu/models/glm-5.3.md) | [`compute/coding-plans/zhipu-coding.json`](../../../compute/coding-plans/zhipu-coding.json) | `contextWindow`、`defaultTemperature`、`defaultTopP`、`maxOutputTokens` | [glm-plan-intl](../../model-sources/providers/zhipu/SOURCES.md#glm-plan-intl)：fetched；[glm53](../../model-sources/providers/zhipu/SOURCES.md#glm53)：fetched | +| [`glm-5.3`](../../model-sources/providers/zhipu/models/glm-5.3.md) | [`compute/model-specs/zhipu.json`](../../../compute/model-specs/zhipu.json) | `spec.contextWindow`、`spec.defaultTemperature`、`spec.defaultTopP`、`spec.maxOutputTokens`、`spec.supportsReasoning` | [glm-plan](../../model-sources/providers/zhipu/SOURCES.md#glm-plan)：fetched；[glm53](../../model-sources/providers/zhipu/SOURCES.md#glm53)：fetched | ## 历史身份/参数不证明当前可用性（0 条） diff --git a/manifest.json b/manifest.json index d3f241b..082a90e 100644 --- a/manifest.json +++ b/manifest.json @@ -1,6 +1,6 @@ { "version": "1.0.0", - "presetDataVersion": 129, + "presetDataVersion": 130, "updatedAt": "2026-10-05", "description": "DesireCore 官方配置中心" }