Files
Yige 647d6f0cfd docs: split model source records and unify project instructions (#113)
* docs: split model sources and consolidate project instructions

* docs: normalize empty model navigation footer
2026-10-03 00:29:46 -07:00

5.4 KiB
Raw Permalink Blame History

gpt-6-astra:模型来源与接入记录

供应商索引 · 官网证据目录

  • 精确模型 ID:gpt-6-astra。
  • 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。

适用接入面

字段级来源记录

各接入面分开记录;价格为 inputPrice/outputPrice 快照,具体单位和限制看配置及引用官网。partial 只确认已列字段,不代表全部参数、平台可用性或账号调用已验收。

compute/providers/openai-codex.json

查看配置 · 接入面说明

当前配置详情(不是全部官网确认值;routing 是本仓策略):

{
  "contextWindow": 1050000,
  "maxOutputTokens": 128000,
  "serviceType": [
    "chat"
  ],
  "capabilities": [
    "chat",
    "reasoning",
    "code",
    "vision",
    "long_context",
    "tool_use",
    "agent"
  ],
  "extra.reasoning": {
    "supportedEfforts": [
      "low",
      "medium",
      "high",
      "xhigh",
      "max"
    ],
    "defaultEffort": "medium"
  }
}
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
gpt-6-astra 1050000 / 128000 USD:未声明 / 未声明 modelName→openai-models openai-models partial 8e958e8a76f90b126b2a3aed30fb7abb2cd2ae16d9df51fa3238518581dcc787

compute/providers/openai.json

查看配置 · 接入面说明

当前配置详情(不是全部官网确认值;routing 是本仓策略):

{
  "contextWindow": 1050000,
  "maxOutputTokens": 128000,
  "serviceType": [
    "chat",
    "responses"
  ],
  "capabilities": [
    "chat",
    "reasoning",
    "code",
    "vision",
    "ultra_long_context",
    "tool_use",
    "agent"
  ],
  "inputPrice": 10,
  "outputPrice": 50,
  "extra.reasoning": {
    "supportedEfforts": [
      "low",
      "medium",
      "high",
      "xhigh",
      "max"
    ],
    "defaultEffort": "medium"
  },
  "extra.cachedInputPrice": 1,
  "extra.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."
}
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
gpt-6-astra 1050000 / 128000 USD:10 / 50 modelName→openai-models openai-models partial bcb8c3cc3c3cb4cb972f27881fc2fc7a54e0dc04c36c827f3e85ddecdb98cacc

compute/model-specs/openai.json

查看配置 · 接入面说明

当前配置详情(不是全部官网确认值;routing 是本仓策略):

{
  "spec.contextWindow": 1050000,
  "spec.maxOutputTokens": 128000,
  "spec.serviceType": [
    "chat"
  ],
  "spec.capabilities": [
    "chat",
    "reasoning",
    "code",
    "vision",
    "long_context",
    "tool_use",
    "agent"
  ],
  "spec.supportsReasoning": true,
  "spec.extra.thinkingOnly": true,
  "routing.tier": "flagship",
  "routing.routingPriority": 10,
  "routing.eligibleForAgent": true,
  "routing.reasoning": {
    "supportedModes": [
      "auto",
      "low",
      "medium",
      "high",
      "xhigh",
      "max"
    ],
    "defaultMode": "medium"
  }
}
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
gpt-6-astra 1050000 / 128000 非计价主数据 id→openai-models openai-models partial 11483d9d9822c28148540e5b3aac06614fc2fec0ceb63c16ecd3b8c22e7adfdf

下次更新核查

  • 对照官网核查精确 ID/别名、是否仍列出、上下文/输入/输出限制及单位。
  • 分别核查推理档位、默认值、是否可关闭思考、采样和强制工具选择限制、多模态输入/输出。
  • 按本接入面核对价格、缓存、阶梯/峰谷、套餐支持;不能把其他平台参数直接复制。
  • 只更新真正复核的字段与引用来源;保留其余待核实项及历史记录。
  • 同步对应 canonical JSON、回归和模型指纹,再运行来源校验。
{
  "formatVersion": 1,
  "supplier": "openai",
  "checkedAt": "2026-10-03",
  "sourceCatalog": "../SOURCES.md"
}