docs: split model source records and unify project instructions (#113)

* docs: split model sources and consolidate project instructions * docs: normalize empty model navigation footer
2026-10-21 05:44:26 +08:00 · 2026-10-03 00:29:46 -07:00
parent cccf14f72d
commit 647d6f0cfd
517 changed files with 39351 additions and 3328 deletions
--- a/docs/model-sources/providers/openai/models/gpt-5.4-mini.md
+++ b/docs/model-sources/providers/openai/models/gpt-5.4-mini.md
@@ -0,0 +1,154 @@
+# gpt-5.4-mini：模型来源与接入记录
+
+[供应商索引](../README.md) · [官网证据目录](../SOURCES.md)
+
+- 精确模型 ID：`gpt-5.4-mini`。
+- 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
+
+- [compute/providers/openai-codex.json](../access/providers--openai-codex.md)
+- [compute/providers/openai.json](../access/providers--openai.md)
+- [compute/model-specs/openai.json](../access/model-specs--openai.md)
+
+## 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](../../../../../compute/providers/openai-codex.json) · [接入面说明](../access/providers--openai-codex.md)
+
+当前配置详情（不是全部官网确认值；routing 是本仓策略）：
+
+<!-- source-details: {"config":"compute/providers/openai-codex.json","id":"gpt-5.4-mini"} -->
+```json
+{
+  "contextWindow": 272000,
+  "maxOutputTokens": 128000,
+  "serviceType": [
+    "chat"
+  ],
+  "capabilities": [
+    "chat",
+    "reasoning",
+    "code",
+    "vision",
+    "long_context",
+    "tool_use",
+    "fast"
+  ]
+}
+```
+<!-- source-details:end -->
+
+<!-- source-config: compute/providers/openai-codex.json -->
+
+| 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-5.4-mini` | 272000 / 128000 | USD：未声明 / 未声明 | 待核实 | [openai-models](../SOURCES.md#openai-models) | pending | `021e6f647c868b9c5385aad2a8de08e14c3bc2b900fd2664e3af4b41ee8af386` |
+
+### compute/providers/openai.json
+
+[查看配置](../../../../../compute/providers/openai.json) · [接入面说明](../access/providers--openai.md)
+
+当前配置详情（不是全部官网确认值；routing 是本仓策略）：
+
+<!-- source-details: {"config":"compute/providers/openai.json","id":"gpt-5.4-mini"} -->
+```json
+{
+  "contextWindow": 400000,
+  "maxOutputTokens": 128000,
+  "serviceType": [
+    "chat",
+    "fast",
+    "responses"
+  ],
+  "capabilities": [
+    "chat",
+    "reasoning",
+    "code",
+    "vision",
+    "long_context",
+    "tool_use",
+    "fast"
+  ],
+  "inputPrice": 0.75,
+  "outputPrice": 4.5,
+  "extra.cachedInputPrice": 0.075
+}
+```
+<!-- source-details:end -->
+
+<!-- source-config: compute/providers/openai.json -->
+
+| 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-5.4-mini` | 400000 / 128000 | USD：0.75 / 4.5 | 待核实 | [openai-models](../SOURCES.md#openai-models) | pending | `b14b746d46d9a0e03d4998ebdb31644d3620121d265bd657e3b9447fc0349dde` |
+
+### compute/model-specs/openai.json
+
+[查看配置](../../../../../compute/model-specs/openai.json) · [接入面说明](../access/model-specs--openai.md)
+
+当前配置详情（不是全部官网确认值；routing 是本仓策略）：
+
+<!-- source-details: {"config":"compute/model-specs/openai.json","id":"gpt-5.4-mini"} -->
+```json
+{
+  "spec.contextWindow": 400000,
+  "spec.maxOutputTokens": 128000,
+  "spec.serviceType": [
+    "chat"
+  ],
+  "spec.capabilities": [
+    "chat",
+    "reasoning",
+    "code",
+    "vision",
+    "long_context",
+    "tool_use",
+    "fast"
+  ],
+  "spec.supportsReasoning": true,
+  "routing.tier": "lightweight",
+  "routing.routingPriority": 25,
+  "routing.eligibleForAgent": true,
+  "routing.reasoning": {
+    "supportedModes": [
+      "auto",
+      "off",
+      "low",
+      "medium",
+      "high",
+      "xhigh"
+    ],
+    "defaultMode": "low"
+  }
+}
+```
+<!-- source-details:end -->
+
+<!-- source-config: compute/model-specs/openai.json -->
+
+| 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-5.4-mini` | 400000 / 128000 | 非计价主数据 | 待核实 | [openai-models](../SOURCES.md#openai-models) | pending | `fea5519f807fe1a18b9ee8f79d556da357f78b8c00f2898b6ea5cf1b7fcc72d5` |
+
+## 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。
+
+<!-- source-metadata:start -->
+```json
+{
+  "formatVersion": 1,
+  "supplier": "openai",
+  "checkedAt": "2026-10-03",
+  "sourceCatalog": "../SOURCES.md"
+}
+```
+<!-- source-metadata:end -->