支持 OpenAI Chat 兼容协议 (#99)

* feat: 完成 openai 接口兼容 * feat: 完成 openai 协议兼容 * fix: 修复测试用例
2026-06-15 12:55:51 +00:00 · 2026-04-03 23:33:17 +08:00
parent 465e9f01c6
commit 00b044e8b2
22 changed files with 2283 additions and 18 deletions
--- a/.github/workflows/ci.yml
+++ b/.github/workflows/ci.yml
@@ -8,7 +8,7 @@ on:

 jobs:
  ci:
-    runs-on: ubuntu-latest
+    runs-on: macos-latest

    steps:
      - uses: actions/checkout@v4
--- a/DEV-LOG.md
+++ b/DEV-LOG.md
@@ -1,5 +1,54 @@
 # DEV-LOG

+## OpenAI 接口兼容 (2026-04-03)
+
+**分支**: `feature/openai`
+
+在 `/login` 流程中新增 "OpenAI Compatible" 选项，支持 Ollama、DeepSeek、vLLM、One API 等兼容 OpenAI Chat Completions API 的第三方服务。用户通过 `/login` 配置后，所有 API 请求自动走 OpenAI 路径。
+
+**改动文件（10 个，+384 / -134）：**
+
+| 文件 | 变更 |
+|------|------|
+| `.github/workflows/ci.yml` | CI runner 从 `ubuntu-latest` 改为 `macos-latest` |
+| `README.md` | TODO 列表新增 "OpenAI 接口兼容" 条目 |
+| `src/components/ConsoleOAuthFlow.tsx` | 新增 `openai_chat_api` OAuth state（含 Base URL / API Key / 3 个模型映射字段）；idle 选择列表新增 "OpenAI Compatible" 选项；完整表单 UI（Tab 切换、Enter 保存）；保存时写入 `modelType: 'openai'` + env 到 settings.json；OAuth 登录时重置 `modelType` 为 `anthropic` |
+| `src/services/api/openai/index.ts` | 从直接 `yield* adaptOpenAIStreamToAnthropic()` 改为完整流处理循环：累积 content blocks（text/tool_use/thinking）、按 `content_block_stop` yield `AssistantMessage`、同时 yield `StreamEvent` 用于实时显示；错误处理改用新签名 `createAssistantAPIErrorMessage({ content, apiError, error })` |
+| `src/services/api/openai/convertMessages.ts` | 输入类型从 Anthropic SDK `BetaMessageParam[]` 改为内部 `(UserMessage \| AssistantMessage)[]`；通过 `msg.type` 而非 `msg.role` 判断角色；从 `msg.message.content` 读取内容；跳过 `cache_edits` / `server_tool_use` 等内部 block 类型 |
+| `src/services/api/openai/modelMapping.ts` | 移除 `OPENAI_MODEL_MAP` JSON 环境变量 + 缓存机制；新增 `getModelFamily()` 按 haiku/sonnet/opus 分类；解析优先级改为：`OPENAI_MODEL` → `ANTHROPIC_DEFAULT_{FAMILY}_MODEL` → `DEFAULT_MODEL_MAP` → 原名透传 |
+| `src/services/api/openai/__tests__/convertMessages.test.ts` | 测试输入从裸 `{ role, content }` 改为 `makeUserMsg()` / `makeAssistantMsg()` 包装的内部格式 |
+| `src/services/api/openai/__tests__/modelMapping.test.ts` | 测试从 `OPENAI_MODEL_MAP` 改为 `ANTHROPIC_DEFAULT_{HAIKU,SONNET,OPUS}_MODEL`；新增 3 个 env var override 测试 |
+| `src/utils/model/providers.ts` | `getAPIProvider()` 新增最高优先级：从 settings.json `modelType` 字段判断；环境变量 `CLAUDE_CODE_USE_OPENAI` 降为次优先 |
+| `src/utils/settings/types.ts` | `SettingsSchema` 新增 `modelType` 字段：`z.enum(['anthropic', 'openai']).optional()` |
+
+**关键设计决策：**
+
+1. **`modelType` 存入 settings.json** — 而非纯环境变量，使 `/login` 配置持久化，重启后仍然生效
+2. **复用 `ANTHROPIC_DEFAULT_*_MODEL` 环境变量** — 而非新增 `OPENAI_MODEL_MAP`，与 Custom Platform 共用同一套模型映射配置，减少用户认知负担
+3. **流处理双 yield** — 同时 yield `AssistantMessage`（给消费方处理工具调用）和 `StreamEvent`（给 REPL 实时渲染），与 Anthropic 路径行为对齐
+4. **OAuth 登录重置 modelType** — 用户切换回官方 Anthropic 登录时自动重置为 `anthropic`，避免残留配置导致请求走错误路径
+
+**配置方式：**
+
+```
+/login → 选择 "OpenAI Compatible" → 填写 Base URL / API Key / 模型名称
+```
+
+或手动编辑 `~/.claude/settings.json`：
+
+```json
+{
+  "modelType": "openai",
+  "env": {
+    "OPENAI_BASE_URL": "http://localhost:11434/v1",
+    "OPENAI_API_KEY": "ollama",
+    "ANTHROPIC_DEFAULT_SONNET_MODEL": "qwen3:32b"
+  }
+}
+```
+
+---
+
 ## Enable Remote Control / BRIDGE_MODE (2026-04-03)

 **PR**: [claude-code-best/claude-code#60](https://github.com/claude-code-best/claude-code/pull/60)
--- a/README.md
+++ b/README.md
@@ -36,6 +36,7 @@
  - [x] 添加自定义 GrowthBook 支持 (GB 也是开源的, 现在你可以配置一个自定义的遥控平台) [文档](https://ccb.agent-aura.top/docs/internals/growthbook-adapter)
  - [x] 自定义 login 模式, 大家可以用这个配置 Claude 的模型!
  - [x] 修复搜索工具的 rg 缺失问题(需要重新 bun i)
+  - [ ] OpenAI 接口兼容! /login 然后配置 OpenAI 平台即可!
 - [ ] V6 大规模重构石山代码, 全面模块分包
  - [ ] V6 将会为全新分支, 届时 main 分支将会封存为历史版本

--- a/bun.lock
+++ b/bun.lock
@@ -89,6 +89,7 @@
        "lru-cache": "^11.2.7",
        "marked": "^17.0.5",
        "modifiers-napi": "workspace:*",
+        "openai": "^4.73.0",
        "p-map": "^7.0.4",
        "picomatch": "^4.0.4",
        "plist": "^3.1.0",
@@ -741,6 +742,8 @@

    "@types/node": ["@types/node@25.5.0", "https://registry.npmmirror.com/@types/node/-/node-25.5.0.tgz", { "dependencies": { "undici-types": "~7.18.0" } }, "sha512-jp2P3tQMSxWugkCUKLRPVUpGaL5MVFwF8RDuSRztfwgN1wmqJeMSbKlnEtQqU8UrhTmzEmZdu2I6v2dpp7XIxw=="],

+    "@types/node-fetch": ["@types/node-fetch@2.6.13", "https://registry.npmmirror.com/@types/node-fetch/-/node-fetch-2.6.13.tgz", { "dependencies": { "@types/node": "*", "form-data": "^4.0.4" } }, "sha512-QGpRVpzSaUs30JBSGPjOg4Uveu384erbHBoT1zeONvyCfwQxIkUshLAOqN/k9EjGviPRmWTTe6aH2qySWKTVSw=="],
+
    "@types/pg": ["@types/pg@8.15.6", "https://registry.npmmirror.com/@types/pg/-/pg-8.15.6.tgz", { "dependencies": { "@types/node": "*", "pg-protocol": "*", "pg-types": "^2.2.0" } }, "sha512-NoaMtzhxOrubeL/7UZuNTrejB4MPAJ0RpxZqXQf2qXuVlTPuG6Y8p4u9dKRaue4yjmC7ZhzVO2/Yyyn25znrPQ=="],

    "@types/pg-pool": ["@types/pg-pool@2.0.7", "https://registry.npmmirror.com/@types/pg-pool/-/pg-pool-2.0.7.tgz", { "dependencies": { "@types/pg": "*" } }, "sha512-U4CwmGVQcbEuqpyju8/ptOKg6gEC+Tqsvj2xS9o1g71bUh8twxnC6ZL5rZKCsGN0iyH0CwgUyc9VR5owNQF9Ng=="],
@@ -763,6 +766,8 @@

    "@xmldom/xmldom": ["@xmldom/xmldom@0.8.12", "https://registry.npmmirror.com/@xmldom/xmldom/-/xmldom-0.8.12.tgz", {}, "sha512-9k/gHF6n/pAi/9tqr3m3aqkuiNosYTurLLUtc7xQ9sxB/wm7WPygCv8GYa6mS0fLJEHhqMC1ATYhz++U/lRHqg=="],

+    "abort-controller": ["abort-controller@3.0.0", "https://registry.npmmirror.com/abort-controller/-/abort-controller-3.0.0.tgz", { "dependencies": { "event-target-shim": "^5.0.0" } }, "sha512-h8lQ8tacZYnR3vNQTgibj+tODHI5/+l06Au2Pcriv/Gmet0eaj4TwWH41sO9wnHDiQsEj19q0drzdWdeAHtweg=="],
+
    "accepts": ["accepts@2.0.0", "https://registry.npmmirror.com/accepts/-/accepts-2.0.0.tgz", { "dependencies": { "mime-types": "^3.0.0", "negotiator": "^1.0.0" } }, "sha512-5cvg6CtKwfgdmVqY1WIiXKc3Q1bkRqGLi+2W/6ao+6Y7gu/RCwRuAhGEzh5B4KlszSuTLgZYuqFqo5bImjNKng=="],

    "acorn": ["acorn@8.16.0", "https://registry.npmmirror.com/acorn/-/acorn-8.16.0.tgz", { "bin": { "acorn": "bin/acorn" } }, "sha512-UVJyE9MttOsBQIDKw1skb9nAwQuR5wuGD3+82K6JgJlm/Y+KI92oNsMNGZCYdDsVtRHSak0pcV5Dno5+4jh9sw=="],
@@ -771,6 +776,8 @@

    "agent-base": ["agent-base@8.0.0", "https://registry.npmmirror.com/agent-base/-/agent-base-8.0.0.tgz", {}, "sha512-QT8i0hCz6C/KQ+KTAbSNwCHDGdmUJl2tp2ZpNlGSWCfhUNVbYG2WLE3MdZGBAgXPV4GAvjGMxo+C1hroyxmZEg=="],

+    "agentkeepalive": ["agentkeepalive@4.6.0", "https://registry.npmmirror.com/agentkeepalive/-/agentkeepalive-4.6.0.tgz", { "dependencies": { "humanize-ms": "^1.2.1" } }, "sha512-kja8j7PjmncONqaTsB8fQ+wE2mSU2DJ9D4XKoJ5PFWIdRMa6SLSN1ff4mOr4jCbfRSsxR4keIiySJU0N9T5hIQ=="],
+
    "ajv": ["ajv@8.18.0", "https://registry.npmmirror.com/ajv/-/ajv-8.18.0.tgz", { "dependencies": { "fast-deep-equal": "^3.1.3", "fast-uri": "^3.0.1", "json-schema-traverse": "^1.0.0", "require-from-string": "^2.0.2" } }, "sha512-PlXPeEWMXMZ7sPYOHqmDyCJzcfNrUr3fGNKtezX14ykXOEIvyK81d+qydx89KY5O71FKMPaQ2vBfBFI5NHR63A=="],

    "ajv-formats": ["ajv-formats@3.0.1", "https://registry.npmmirror.com/ajv-formats/-/ajv-formats-3.0.1.tgz", { "dependencies": { "ajv": "^8.0.0" } }, "sha512-8iUql50EUR+uUcdRQ3HDqa6EVyo3docL8g5WJ3FNcWmu62IbkGUue/pEyLBW8VGKKucTPgqeks4fIU1DA4yowQ=="],
@@ -927,6 +934,8 @@

    "etag": ["etag@1.8.1", "https://registry.npmmirror.com/etag/-/etag-1.8.1.tgz", {}, "sha512-aIL5Fx7mawVa300al2BnEE4iNvo1qETxLrPI/o05L7z6go7fCw1J6EQmbK4FmJ2AS7kgVF/KEZWufBfdClMcPg=="],

+    "event-target-shim": ["event-target-shim@5.0.1", "https://registry.npmmirror.com/event-target-shim/-/event-target-shim-5.0.1.tgz", {}, "sha512-i/2XbnSz/uxRCU6+NdVJgKWDTM427+MqYbkQzD321DuCQJUqOuJKIA0IM2+W2xtYHdKOmZ4dR6fExsd4SXL+WQ=="],
+
    "eventsource": ["eventsource@3.0.7", "https://registry.npmmirror.com/eventsource/-/eventsource-3.0.7.tgz", { "dependencies": { "eventsource-parser": "^3.0.1" } }, "sha512-CRT1WTyuQoD771GW56XEZFQ/ZoSfWid1alKGDYMmkt2yl8UXrVR4pspqWNEcqKvVIzg6PAltWjxcSSPrboA4iA=="],

    "eventsource-parser": ["eventsource-parser@3.0.6", "https://registry.npmmirror.com/eventsource-parser/-/eventsource-parser-3.0.6.tgz", {}, "sha512-Vo1ab+QXPzZ4tCa8SwIHJFaSzy4R6SHf7BY79rFBDf0idraZWAkYrDjDj8uWaSm3S2TK+hJ7/t1CEmZ7jXw+pg=="],
@@ -973,8 +982,12 @@

    "form-data": ["form-data@4.0.5", "https://registry.npmmirror.com/form-data/-/form-data-4.0.5.tgz", { "dependencies": { "asynckit": "^0.4.0", "combined-stream": "^1.0.8", "es-set-tostringtag": "^2.1.0", "hasown": "^2.0.2", "mime-types": "^2.1.12" } }, "sha512-8RipRLol37bNs2bhoV67fiTEvdTrbMUYcFTiy3+wuuOnUog2QBHCZWXDRijWQfAkhBj2Uf5UnVaiWwA5vdd82w=="],

+    "form-data-encoder": ["form-data-encoder@1.7.2", "https://registry.npmmirror.com/form-data-encoder/-/form-data-encoder-1.7.2.tgz", {}, "sha512-qfqtYan3rxrnCk1VYaA4H+Ms9xdpPqvLZa6xmMgFvhO32x7/3J/ExcTd6qpxM0vH2GdMI+poehyBZvqfMTto8A=="],
+
    "formatly": ["formatly@0.3.0", "https://registry.npmmirror.com/formatly/-/formatly-0.3.0.tgz", { "dependencies": { "fd-package-json": "^2.0.0" }, "bin": { "formatly": "bin/index.mjs" } }, "sha512-9XNj/o4wrRFyhSMJOvsuyMwy8aUfBaZ1VrqHVfohyXf0Sw0e+yfKG+xZaY3arGCOMdwFsqObtzVOc1gU9KiT9w=="],

+    "formdata-node": ["formdata-node@4.4.1", "https://registry.npmmirror.com/formdata-node/-/formdata-node-4.4.1.tgz", { "dependencies": { "node-domexception": "1.0.0", "web-streams-polyfill": "4.0.0-beta.3" } }, "sha512-0iirZp3uVDjVGt9p49aTaqjk84TrglENEDuqfdlZQ1roC9CWlPk6Avf8EEnZNcAqPonwkG35x4n3ww/1THYAeQ=="],
+
    "formdata-polyfill": ["formdata-polyfill@4.0.10", "https://registry.npmmirror.com/formdata-polyfill/-/formdata-polyfill-4.0.10.tgz", { "dependencies": { "fetch-blob": "^3.1.2" } }, "sha512-buewHzMvYL29jdeQTVILecSaZKnt/RJWjoZCF5OW60Z67/GmSLBkOFM7qh1PI3zFNtJbaZL5eQu1vLfazOwj4g=="],

    "forwarded": ["forwarded@0.2.0", "https://registry.npmmirror.com/forwarded/-/forwarded-0.2.0.tgz", {}, "sha512-buRG0fpBtRHSTCOASe6hD258tEubFoRLb4ZNA6NxMVHNw2gOcwHo9wyablzMzOA5z9xA9L1KNjk/Nt6MT9aYow=="],
@@ -1045,6 +1058,8 @@

    "human-signals": ["human-signals@8.0.1", "https://registry.npmmirror.com/human-signals/-/human-signals-8.0.1.tgz", {}, "sha512-eKCa6bwnJhvxj14kZk5NCPc6Hb6BdsU9DZcOnmQKSnO1VKrfV0zCvtttPZUsBvjmNDn8rpcJfpwSYnHBjc95MQ=="],

+    "humanize-ms": ["humanize-ms@1.2.1", "https://registry.npmmirror.com/humanize-ms/-/humanize-ms-1.2.1.tgz", { "dependencies": { "ms": "^2.0.0" } }, "sha512-Fl70vYtsAFb/C06PTS9dZBo7ihau+Tu/DNCk/OyHhea07S+aeMWpFFkUaXRa8fI+ScZbEI8dfSxwY7gxZ9SAVQ=="],
+
    "iconv-lite": ["iconv-lite@0.7.2", "https://registry.npmmirror.com/iconv-lite/-/iconv-lite-0.7.2.tgz", { "dependencies": { "safer-buffer": ">= 2.1.2 < 3.0.0" } }, "sha512-im9DjEDQ55s9fL4EYzOAv0yMqmMBSZp6G0VvFyTMPKWxiSBHUj9NW/qqLmXUwXrrM7AvqSlTCfvqRb0cM8yYqw=="],

    "ignore": ["ignore@7.0.5", "https://registry.npmmirror.com/ignore/-/ignore-7.0.5.tgz", {}, "sha512-Hs59xBNfUIunMFgWAbGX5cq6893IbWg4KnrjbYwX3tx0ztorVgTDA6B2sxf8ejHJ4wz8BqGUMYlnzNBer5NvGg=="],
@@ -1181,7 +1196,7 @@

    "node-domexception": ["node-domexception@1.0.0", "https://registry.npmmirror.com/node-domexception/-/node-domexception-1.0.0.tgz", {}, "sha512-/jKZoMpw0F8GRwl4/eLROPA3cfcXtLApP0QzLmUT/HuPCZWyB7IY9ZrMeKw2O/nFIqPQB3PVM9aYm0F312AXDQ=="],

-    "node-fetch": ["node-fetch@3.3.2", "https://registry.npmmirror.com/node-fetch/-/node-fetch-3.3.2.tgz", { "dependencies": { "data-uri-to-buffer": "^4.0.0", "fetch-blob": "^3.1.4", "formdata-polyfill": "^4.0.10" } }, "sha512-dRB78srN/l6gqWulah9SrxeYnxeddIG30+GOqK/9OlLVyLg3HPnr6SqOWTWOXKRwC2eGYCkZ59NNuSgvSrpgOA=="],
+    "node-fetch": ["node-fetch@2.7.0", "https://registry.npmmirror.com/node-fetch/-/node-fetch-2.7.0.tgz", { "dependencies": { "whatwg-url": "^5.0.0" }, "peerDependencies": { "encoding": "^0.1.0" }, "optionalPeers": ["encoding"] }, "sha512-c4FRfUm/dbcWZ7U+1Wq0AwCyFL+3nt2bEw05wfxSz+DWpWsitgmSgYmy2dQdWyKC1694ELPqMs/YzUSNozLt8A=="],

    "node-forge": ["node-forge@1.4.0", "https://registry.npmmirror.com/node-forge/-/node-forge-1.4.0.tgz", {}, "sha512-LarFH0+6VfriEhqMMcLX2F7SwSXeWwnEAJEsYm5QKWchiVYVvJyV9v7UDvUv+w5HO23ZpQTXDv/GxdDdMyOuoQ=="],

@@ -1197,6 +1212,8 @@

    "open": ["open@10.2.0", "https://registry.npmmirror.com/open/-/open-10.2.0.tgz", { "dependencies": { "default-browser": "^5.2.1", "define-lazy-prop": "^3.0.0", "is-inside-container": "^1.0.0", "wsl-utils": "^0.1.0" } }, "sha512-YgBpdJHPyQ2UE5x+hlSXcnejzAvD0b22U2OuAP+8OnlJT+PjWPxtgmGqKKc+RgTM63U9gN0YzrYc71R2WT/hTA=="],

+    "openai": ["openai@4.104.0", "https://registry.npmmirror.com/openai/-/openai-4.104.0.tgz", { "dependencies": { "@types/node": "^18.11.18", "@types/node-fetch": "^2.6.4", "abort-controller": "^3.0.0", "agentkeepalive": "^4.2.1", "form-data-encoder": "1.7.2", "formdata-node": "^4.3.2", "node-fetch": "^2.6.7" }, "peerDependencies": { "ws": "^8.18.0", "zod": "^3.23.8" }, "optionalPeers": ["ws", "zod"], "bin": { "openai": "bin/cli" } }, "sha512-p99EFNsA/yX6UhVO93f5kJsDRLAg+CTA2RBqdHK4RtK8u5IJw32Hyb2dTGKbnnFmnuoBv5r7Z2CURI9sGZpSuA=="],
+
    "os-tmpdir": ["os-tmpdir@1.0.2", "https://registry.npmmirror.com/os-tmpdir/-/os-tmpdir-1.0.2.tgz", {}, "sha512-D2FR03Vir7FIu45XBY20mTb+/ZSWB00sjU9jdQXt83gDrI4Ztz5Fs7/yy74g2N5SVQY4xY1qDr4rNddwYRVX0g=="],

    "oxc-parser": ["oxc-parser@0.121.0", "https://registry.npmmirror.com/oxc-parser/-/oxc-parser-0.121.0.tgz", { "dependencies": { "@oxc-project/types": "^0.121.0" }, "optionalDependencies": { "@oxc-parser/binding-android-arm-eabi": "0.121.0", "@oxc-parser/binding-android-arm64": "0.121.0", "@oxc-parser/binding-darwin-arm64": "0.121.0", "@oxc-parser/binding-darwin-x64": "0.121.0", "@oxc-parser/binding-freebsd-x64": "0.121.0", "@oxc-parser/binding-linux-arm-gnueabihf": "0.121.0", "@oxc-parser/binding-linux-arm-musleabihf": "0.121.0", "@oxc-parser/binding-linux-arm64-gnu": "0.121.0", "@oxc-parser/binding-linux-arm64-musl": "0.121.0", "@oxc-parser/binding-linux-ppc64-gnu": "0.121.0", "@oxc-parser/binding-linux-riscv64-gnu": "0.121.0", "@oxc-parser/binding-linux-riscv64-musl": "0.121.0", "@oxc-parser/binding-linux-s390x-gnu": "0.121.0", "@oxc-parser/binding-linux-x64-gnu": "0.121.0", "@oxc-parser/binding-linux-x64-musl": "0.121.0", "@oxc-parser/binding-openharmony-arm64": "0.121.0", "@oxc-parser/binding-wasm32-wasi": "0.121.0", "@oxc-parser/binding-win32-arm64-msvc": "0.121.0", "@oxc-parser/binding-win32-ia32-msvc": "0.121.0", "@oxc-parser/binding-win32-x64-msvc": "0.121.0" } }, "sha512-ek9o58+SCv6AV7nchiAcUJy1DNE2CC5WRdBcO0mF+W4oRjNQfPO7b3pLjTHSFECpHkKGOZSQxx3hk8viIL5YCg=="],
@@ -1417,7 +1434,7 @@

    "walk-up-path": ["walk-up-path@4.0.0", "https://registry.npmmirror.com/walk-up-path/-/walk-up-path-4.0.0.tgz", {}, "sha512-3hu+tD8YzSLGuFYtPRb48vdhKMi0KQV5sn+uWr8+7dMEq/2G/dtLrdDinkLjqq5TIbIBjYJ4Ax/n3YiaW7QM8A=="],

-    "web-streams-polyfill": ["web-streams-polyfill@3.3.3", "https://registry.npmmirror.com/web-streams-polyfill/-/web-streams-polyfill-3.3.3.tgz", {}, "sha512-d2JWLCivmZYTSIoge9MsgFCZrt571BikcWGYkjC1khllbTeDlGqZ2D8vD8E/lJa8WGWbb7Plm8/XJYV7IJHZZw=="],
+    "web-streams-polyfill": ["web-streams-polyfill@4.0.0-beta.3", "https://registry.npmmirror.com/web-streams-polyfill/-/web-streams-polyfill-4.0.0-beta.3.tgz", {}, "sha512-QW95TCTaHmsYfHDybGMwO5IJIM93I/6vTRk+daHTWFPhwh+C8Cg7j7XyKrwrj8Ib6vYXe0ocYNrmzY4xAAN6ug=="],

    "webidl-conversions": ["webidl-conversions@3.0.1", "https://registry.npmmirror.com/webidl-conversions/-/webidl-conversions-3.0.1.tgz", {}, "sha512-2JAn3z8AR6rjK8Sm8orRC0h/bcl/DqL7tRPdGZ4I1CjdF+EaMLmYxBHyXuKL849eucPFhvBoxMsflfOb8kxaeQ=="],

@@ -1755,10 +1772,14 @@

    "external-editor/iconv-lite": ["iconv-lite@0.4.24", "https://registry.npmmirror.com/iconv-lite/-/iconv-lite-0.4.24.tgz", { "dependencies": { "safer-buffer": ">= 2.1.2 < 3" } }, "sha512-v3MXnZAcvnywkTUEZomIActle7RXXeedOR31wwl7VlyoXO4Qi9arvSenNQWne1TcRwhCL1HwLI21bEqdpj8/rA=="],

+    "fetch-blob/web-streams-polyfill": ["web-streams-polyfill@3.3.3", "https://registry.npmmirror.com/web-streams-polyfill/-/web-streams-polyfill-3.3.3.tgz", {}, "sha512-d2JWLCivmZYTSIoge9MsgFCZrt571BikcWGYkjC1khllbTeDlGqZ2D8vD8E/lJa8WGWbb7Plm8/XJYV7IJHZZw=="],
+
    "form-data/mime-types": ["mime-types@2.1.35", "https://registry.npmmirror.com/mime-types/-/mime-types-2.1.35.tgz", { "dependencies": { "mime-db": "1.52.0" } }, "sha512-ZDY+bPm5zTTF+YpCrAU9nK0UgICYPT0QtT1NZWFv4s++TNkcgVaT0g6+4R2uI4MjQjzysHB1zxuWL50hzaeXiw=="],

    "gaxios/https-proxy-agent": ["https-proxy-agent@7.0.6", "https://registry.npmmirror.com/https-proxy-agent/-/https-proxy-agent-7.0.6.tgz", { "dependencies": { "agent-base": "^7.1.2", "debug": "4" } }, "sha512-vK9P5/iUfdl95AI+JVyUuIcVtd4ofvtrOr3HNtM2yxC9bnMbEdp3x01OhQNnjb8IJYi38VlTE3mBXwcfvywuSw=="],

+    "gaxios/node-fetch": ["node-fetch@3.3.2", "https://registry.npmmirror.com/node-fetch/-/node-fetch-3.3.2.tgz", { "dependencies": { "data-uri-to-buffer": "^4.0.0", "fetch-blob": "^3.1.4", "formdata-polyfill": "^4.0.10" } }, "sha512-dRB78srN/l6gqWulah9SrxeYnxeddIG30+GOqK/9OlLVyLg3HPnr6SqOWTWOXKRwC2eGYCkZ59NNuSgvSrpgOA=="],
+
    "gtoken/gaxios": ["gaxios@6.7.1", "https://registry.npmmirror.com/gaxios/-/gaxios-6.7.1.tgz", { "dependencies": { "extend": "^3.0.2", "https-proxy-agent": "^7.0.1", "is-stream": "^2.0.0", "node-fetch": "^2.6.9", "uuid": "^9.0.1" } }, "sha512-LDODD4TMYx7XXdpwxAVRAIAuB0bzv0s+ywFonY46k126qzQHT9ygyoa9tncmOiQmmDrik65UYsEkv3lbfqQ3yQ=="],

    "http-proxy-agent/agent-base": ["agent-base@7.1.4", "https://registry.npmmirror.com/agent-base/-/agent-base-7.1.4.tgz", {}, "sha512-MnA+YT8fwfJPgBx3m60MNqakm30XOkyIoH1y6huTQvC0PwZG7ki8NacLBcrPbNoo8vEZy7Jpuk7+jMO+CUovTQ=="],
@@ -1773,6 +1794,8 @@

    "npm-run-path/path-key": ["path-key@4.0.0", "https://registry.npmmirror.com/path-key/-/path-key-4.0.0.tgz", {}, "sha512-haREypq7xkM7ErfgIyA0z+Bj4AGKlMSdlQE2jvJo6huWD1EdkKYV+G/T4nq0YEF2vgTT8kqMFKo1uHn950r4SQ=="],

+    "openai/@types/node": ["@types/node@18.19.130", "https://registry.npmmirror.com/@types/node/-/node-18.19.130.tgz", { "dependencies": { "undici-types": "~5.26.4" } }, "sha512-GRaXQx6jGfL8sKfaIDD6OupbIHBr9jv7Jnaml9tB7l4v068PAOXqfcujMMo5PhbIs6ggR1XODELqahT2R8v0fg=="],
+
    "parse5-htmlparser2-tree-adapter/parse5": ["parse5@6.0.1", "https://registry.npmmirror.com/parse5/-/parse5-6.0.1.tgz", {}, "sha512-Ofn/CTFzRGTTxwpNEs9PP93gXShHcTq255nzRYSKe8AkVpZY7e1fpmTfOyoIvjP5HG7Z2ZM7VS9PPhQGW2pOpw=="],

    "proper-lockfile/signal-exit": ["signal-exit@3.0.7", "https://registry.npmmirror.com/signal-exit/-/signal-exit-3.0.7.tgz", {}, "sha512-wnD2ZE+l+SPC/uoS0vXeE9L1+0wuaMqKlfz9AMUo38JsyLSBWSFcHR1Rri62LZc12vLr1gb3jl7iwQhgwpAbGQ=="],
@@ -1909,8 +1932,6 @@

    "gtoken/gaxios/is-stream": ["is-stream@2.0.1", "https://registry.npmmirror.com/is-stream/-/is-stream-2.0.1.tgz", {}, "sha512-hFoiJiTl63nn+kstHGBtewWSKnQLpyb155KHheA1l39uvtO9nWIop1p3udqPcUd/xbF1VLMO4n7OI6p7RbngDg=="],

-    "gtoken/gaxios/node-fetch": ["node-fetch@2.7.0", "https://registry.npmmirror.com/node-fetch/-/node-fetch-2.7.0.tgz", { "dependencies": { "whatwg-url": "^5.0.0" }, "peerDependencies": { "encoding": "^0.1.0" }, "optionalPeers": ["encoding"] }, "sha512-c4FRfUm/dbcWZ7U+1Wq0AwCyFL+3nt2bEw05wfxSz+DWpWsitgmSgYmy2dQdWyKC1694ELPqMs/YzUSNozLt8A=="],
-
    "gtoken/gaxios/uuid": ["uuid@9.0.1", "https://registry.npmmirror.com/uuid/-/uuid-9.0.1.tgz", { "bin": { "uuid": "dist/bin/uuid" } }, "sha512-b+1eJOlsR9K8HJpow9Ok3fiWOWSIcIzXodvv0rQjVoOVNpWMpxf1wZNpt4y9h10odCNrqnYp1OBzRktckBe3sA=="],

    "image-processor-napi/sharp/@img/sharp-darwin-arm64": ["@img/sharp-darwin-arm64@0.33.5", "https://registry.npmmirror.com/@img/sharp-darwin-arm64/-/sharp-darwin-arm64-0.33.5.tgz", { "optionalDependencies": { "@img/sharp-libvips-darwin-arm64": "1.0.4" }, "os": "darwin", "cpu": "arm64" }, "sha512-UT4p+iz/2H4twwAoLCqfA9UH5pI6DggwKEGuaPy7nCVQ8ZsiY5PIcrRvD1DzuY3qYL07NtIQcWnBSY/heikIFQ=="],
@@ -1951,6 +1972,8 @@

    "image-processor-napi/sharp/@img/sharp-win32-x64": ["@img/sharp-win32-x64@0.33.5", "https://registry.npmmirror.com/@img/sharp-win32-x64/-/sharp-win32-x64-0.33.5.tgz", { "os": "win32", "cpu": "x64" }, "sha512-MpY/o8/8kj+EcnxwvrP4aTJSWw/aZ7JIGR4aBeZkZw5B7/Jn+tY9/VNwtcoGmdT7GfggGIU4kygOMSbYnOrAbg=="],

+    "openai/@types/node/undici-types": ["undici-types@5.26.5", "https://registry.npmmirror.com/undici-types/-/undici-types-5.26.5.tgz", {}, "sha512-JlCMO+ehdEIKqlFxk6IfVoAUVmgz7cU7zD/h9XZ0qzeosSHmUJVOzSQvvYSYWXkFXC+IfLKSIffhv0sVZup6pA=="],
+
    "qrcode/yargs/cliui": ["cliui@6.0.0", "https://registry.npmmirror.com/cliui/-/cliui-6.0.0.tgz", { "dependencies": { "string-width": "^4.2.0", "strip-ansi": "^6.0.0", "wrap-ansi": "^6.2.0" } }, "sha512-t6wbgtoCXvAzst7QgXxJYqPt0usEfbgQdftEPbLL/cvv6HPE5VgvqCuAIDR0NgU52ds6rFwqrgakNLrHEjCbrQ=="],

    "qrcode/yargs/string-width": ["string-width@4.2.3", "https://registry.npmmirror.com/string-width/-/string-width-4.2.3.tgz", { "dependencies": { "emoji-regex": "^8.0.0", "is-fullwidth-code-point": "^3.0.0", "strip-ansi": "^6.0.1" } }, "sha512-wKyQRQpjJ0sIp62ErSZdGsjMJWsap5oRNihHhu6G7JVO/9jIB6UyevL+tXuOqrng8j/cxKTWyWUwvSTriiZz/g=="],
@@ -1969,8 +1992,6 @@

    "@anthropic-ai/vertex-sdk/google-auth-library/gaxios/is-stream": ["is-stream@2.0.1", "https://registry.npmmirror.com/is-stream/-/is-stream-2.0.1.tgz", {}, "sha512-hFoiJiTl63nn+kstHGBtewWSKnQLpyb155KHheA1l39uvtO9nWIop1p3udqPcUd/xbF1VLMO4n7OI6p7RbngDg=="],

-    "@anthropic-ai/vertex-sdk/google-auth-library/gaxios/node-fetch": ["node-fetch@2.7.0", "https://registry.npmmirror.com/node-fetch/-/node-fetch-2.7.0.tgz", { "dependencies": { "whatwg-url": "^5.0.0" }, "peerDependencies": { "encoding": "^0.1.0" }, "optionalPeers": ["encoding"] }, "sha512-c4FRfUm/dbcWZ7U+1Wq0AwCyFL+3nt2bEw05wfxSz+DWpWsitgmSgYmy2dQdWyKC1694ELPqMs/YzUSNozLt8A=="],
-
    "@anthropic-ai/vertex-sdk/google-auth-library/gaxios/uuid": ["uuid@9.0.1", "https://registry.npmmirror.com/uuid/-/uuid-9.0.1.tgz", { "bin": { "uuid": "dist/bin/uuid" } }, "sha512-b+1eJOlsR9K8HJpow9Ok3fiWOWSIcIzXodvv0rQjVoOVNpWMpxf1wZNpt4y9h10odCNrqnYp1OBzRktckBe3sA=="],

    "@anthropic-ai/vertex-sdk/google-auth-library/gcp-metadata/google-logging-utils": ["google-logging-utils@0.0.2", "https://registry.npmmirror.com/google-logging-utils/-/google-logging-utils-0.0.2.tgz", {}, "sha512-NEgUnEcBiP5HrPzufUkBzJOD/Sxsco3rLNo1F1TNf7ieU8ryUzBhqba8r756CjLX7rn3fHl6iLEwPYuqpoKgQQ=="],
--- a/docs/plans/openai-compatibility.md
+++ b/docs/plans/openai-compatibility.md
@@ -0,0 +1,421 @@
+# OpenAI 协议兼容层
+
+## 概述
+
+claude-code 支持通过 OpenAI Chat Completions API（`/v1/chat/completions`）兼容任意 OpenAI 协议端点，包括 Ollama、DeepSeek、vLLM、One API、LiteLLM 等。
+
+核心策略为**流适配器模式**：在 `queryModel()` 中插入提前返回分支，将 Anthropic 格式请求转为 OpenAI 格式，调用 OpenAI SDK，再将 SSE 流转换回 `BetaRawMessageStreamEvent` 格式。下游代码（流处理循环、query.ts、QueryEngine.ts、REPL）**完全不改**。
+
+## 环境变量
+
+| 变量 | 必需 | 说明 |
+|---|---|---|
+| `CLAUDE_CODE_USE_OPENAI` | 是 | 设为 `1` 启用 OpenAI 后端 |
+| `OPENAI_API_KEY` | 是 | API key（Ollama 等可设为任意值） |
+| `OPENAI_BASE_URL` | 推荐 | 端点 URL（如 `http://localhost:11434/v1`） |
+| `OPENAI_MODEL` | 可选 | 覆盖所有请求的模型名（跳过映射） |
+| `OPENAI_MODEL_MAP` | 可选 | JSON 映射，如 `{"claude-sonnet-4-6":"gpt-4o"}` |
+| `OPENAI_ORG_ID` | 可选 | Organization ID |
+| `OPENAI_PROJECT_ID` | 可选 | Project ID |
+
+### 使用示例
+
+```bash
+# Ollama
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=ollama \
+OPENAI_BASE_URL=http://localhost:11434/v1 \
+OPENAI_MODEL=qwen2.5-coder-32b \
+bun run dev
+
+# DeepSeek（自动支持 Thinking）
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=sk-xxx \
+OPENAI_BASE_URL=https://api.deepseek.com/v1 \
+OPENAI_MODEL=deepseek-chat \
+bun run dev
+
+# vLLM
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=token-abc123 \
+OPENAI_BASE_URL=http://localhost:8000/v1 \
+OPENAI_MODEL=Qwen/Qwen2.5-Coder-32B-Instruct \
+bun run dev
+
+# One API / LiteLLM
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=sk-your-key \
+OPENAI_BASE_URL=https://your-one-api.example.com/v1 \
+OPENAI_MODEL=gpt-4o \
+bun run dev
+
+# 自定义模型映射
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=sk-xxx \
+OPENAI_BASE_URL=https://my-gateway.example.com/v1 \
+OPENAI_MODEL_MAP='{"claude-sonnet-4-6":"gpt-4o-2024-11-20","claude-haiku-4-5":"gpt-4o-mini"}' \
+bun run dev
+```
+
+## 架构
+
+### 请求流程
+
+```
+queryModel() [claude.ts]
+  ├── 共享预处理（消息归一化、工具过滤、媒体裁剪）
+  └── if (getAPIProvider() === 'openai')
+      └── queryModelOpenAI() [openai/index.ts]
+          ├── resolveOpenAIModel()          → 解析模型名
+          ├── normalizeMessagesForAPI()      → 共享消息预处理
+          ├── toolToAPISchema()              → 构建工具 schema
+          ├── anthropicMessagesToOpenAI()    → 消息格式转换
+          ├── anthropicToolsToOpenAI()       → 工具格式转换
+          ├── openai.chat.completions.create({ stream: true })
+          └── adaptOpenAIStreamToAnthropic() → 流格式转换
+              ├── delta.reasoning_content    → thinking 块
+              ├── delta.content             → text 块
+              ├── delta.tool_calls          → tool_use 块
+              ├── usage.cached_tokens       → cache_read_input_tokens
+              └── yield BetaRawMessageStreamEvent
+```
+
+### 模型名解析优先级
+
+`resolveOpenAIModel()` 的解析顺序：
+
+1. `OPENAI_MODEL` 环境变量 → 直接使用，覆盖所有
+2. `OPENAI_MODEL_MAP` JSON 查表 → 自定义映射
+3. 内置默认映射（见下表）
+4. 以上都不匹配 → 原名透传
+
+### 内置模型映射
+
+| Anthropic 模型 | OpenAI 映射 |
+|---|---|
+| `claude-sonnet-4-6` | `gpt-4o` |
+| `claude-sonnet-4-5-20250929` | `gpt-4o` |
+| `claude-sonnet-4-20250514` | `gpt-4o` |
+| `claude-3-7-sonnet-20250219` | `gpt-4o` |
+| `claude-3-5-sonnet-20241022` | `gpt-4o` |
+| `claude-opus-4-6` | `o3` |
+| `claude-opus-4-5-20251101` | `o3` |
+| `claude-opus-4-1-20250805` | `o3` |
+| `claude-opus-4-20250514` | `o3` |
+| `claude-haiku-4-5-20251001` | `gpt-4o-mini` |
+| `claude-3-5-haiku-20241022` | `gpt-4o-mini` |
+
+同时会自动剥离 `[1m]` 后缀（Claude 特有的 modifier）。
+
+## 文件结构
+
+### 新增文件
+
+```
+src/services/api/openai/
+├── client.ts              # OpenAI SDK 客户端工厂（~50 行）
+├── convertMessages.ts     # Anthropic → OpenAI 消息格式转换（~190 行）
+├── convertTools.ts        # Anthropic → OpenAI 工具格式转换（~70 行）
+├── streamAdapter.ts       # SSE 流转换核心，含 thinking + caching（~270 行）
+├── modelMapping.ts        # 模型名解析（~60 行）
+├── index.ts               # 公共入口 queryModelOpenAI()（~110 行）
+└── __tests__/
+    ├── convertMessages.test.ts   # 10 个测试
+    ├── convertTools.test.ts      # 7 个测试
+    ├── modelMapping.test.ts      # 6 个测试
+    └── streamAdapter.test.ts     # 14 个测试（含 thinking + caching）
+```
+
+### 修改文件
+
+| 文件 | 改动 |
+|---|---|
+| `src/utils/model/providers.ts` | 添加 `'openai'` provider 类型 + `CLAUDE_CODE_USE_OPENAI` 检查（最高优先级） |
+| `src/utils/model/configs.ts` | 每个 ModelConfig 添加 `openai` 键 |
+| `src/services/api/claude.ts` | 在 `stripExcessMediaItems()` 后插入 OpenAI 提前返回分支（~8 行） |
+| `package.json` | 添加 `"openai": "^4.73.0"` 依赖 |
+
+## 消息转换规则
+
+### Anthropic → OpenAI
+
+| Anthropic | OpenAI |
+|---|---|
+| `system` prompt（`string[]`） | `role: "system"` 消息（`\n\n` 拼接） |
+| `user` + `text` 块 | `role: "user"` 消息 |
+| `assistant` + `text` 块 | `role: "assistant"` + `content` |
+| `assistant` + `tool_use` 块 | `role: "assistant"` + `tool_calls[]` |
+| `user` + `tool_result` 块 | `role: "tool"` + `tool_call_id` |
+| `thinking` 块 | 静默丢弃（请求侧） |
+
+### 工具转换
+
+| Anthropic | OpenAI |
+|---|---|
+| `{ name, description, input_schema }` | `{ type: "function", function: { name, description, parameters } }` |
+| `cache_control`, `defer_loading` 等字段 | 剥离 |
+| `tool_choice: { type: "auto" }` | `"auto"` |
+| `tool_choice: { type: "any" }` | `"required"` |
+| `tool_choice: { type: "tool", name }` | `{ type: "function", function: { name } }` |
+
+### 消息转换示例
+
+```
+Anthropic:                              OpenAI:
+[
+  system: ["You are helpful."],         [
+                                          { role: "system",
+  { role: "user",                          content: "You are helpful." },
+    content: [                            { role: "user",
+      { type: "text", text: "Run ls" }      content: "Run ls"
+    ]                                     },
+  },                                      { role: "assistant",
+  { role: "assistant",                     content: "I'll check.",
+    content: [                            tool_calls: [{
+      { type: "text", text: "I'll check."},  id: "tu_123",
+      { type: "tool_use",                    type: "function",
+        id: "tu_123", name: "bash",          function: {
+        input: { command: "ls" } }             name: "bash",
+    ]                                           arguments: '{"command":"ls"}'
+  },                                      }] }
+  { role: "user",                        { role: "tool",
+    content: [                              tool_call_id: "tu_123",
+      { type: "tool_result",                content: "file1\nfile2"
+        tool_use_id: "tu_123",            }
+        content: "file1\nfile2"          ]
+    ]
+  }
+]
+```
+
+## 流转换规则
+
+### SSE Chunk → Anthropic Event 映射
+
+| OpenAI Chunk | Anthropic Event |
+|---|---|
+| 首个 chunk | `message_start`（含 usage） |
+| `delta.reasoning_content` | `content_block_start(thinking)` + `thinking_delta` |
+| `delta.content` | `content_block_start(text)` + `text_delta` |
+| `delta.tool_calls` | `content_block_start(tool_use)` + `input_json_delta` |
+| `finish_reason: "stop"` | `message_delta(stop_reason: "end_turn")` |
+| `finish_reason: "tool_calls"` | `message_delta(stop_reason: "tool_use")` |
+| `finish_reason: "length"` | `message_delta(stop_reason: "max_tokens")` |
+
+### 块顺序
+
+当模型返回 `reasoning_content` 时（如 DeepSeek），块顺序与 Anthropic 一致：
+
+```
+thinking block (index 0)  ← delta.reasoning_content
+text block    (index 1)   ← delta.content
+```
+
+或：
+
+```
+thinking block (index 0)  ← delta.reasoning_content
+tool_use block (index 1)  ← delta.tool_calls
+```
+
+无 `reasoning_content` 时：
+
+```
+text block    (index 0)   ← delta.content
+tool_use block (index 1)  ← delta.tool_calls（如果有）
+```
+
+### finish_reason 映射
+
+| OpenAI | Anthropic |
+|---|---|
+| `stop` | `end_turn` |
+| `tool_calls` | `tool_use` |
+| `length` | `max_tokens` |
+| `content_filter` | `end_turn` |
+
+### 事件序列示例
+
+**纯文本响应**：
+```
+OpenAI chunks:
+  delta.content = "Hello"
+  delta.content = " world"
+  finish_reason = "stop"
+
+→ Anthropic events:
+  message_start       { message: { id, role: 'assistant', usage: {...} } }
+  content_block_start { index: 0, content_block: { type: 'text' } }
+  content_block_delta { index: 0, delta: { type: 'text_delta', text: 'Hello' } }
+  content_block_delta { index: 0, delta: { type: 'text_delta', text: ' world' } }
+  content_block_stop  { index: 0 }
+  message_delta       { delta: { stop_reason: 'end_turn' } }
+  message_stop
+```
+
+**Thinking + 文本（DeepSeek 风格）**：
+```
+OpenAI chunks:
+  delta.reasoning_content = "Let me think..."
+  delta.reasoning_content = " step by step."
+  delta.content = "The answer is 42."
+  finish_reason = "stop"
+
+→ Anthropic events:
+  message_start       { ... }
+  content_block_start { index: 0, content_block: { type: 'thinking', signature: '' } }
+  content_block_delta { index: 0, delta: { type: 'thinking_delta', thinking: 'Let me think...' } }
+  content_block_delta { index: 0, delta: { type: 'thinking_delta', thinking: ' step by step.' } }
+  content_block_stop  { index: 0 }
+  content_block_start { index: 1, content_block: { type: 'text' } }
+  content_block_delta { index: 1, delta: { type: 'text_delta', text: 'The answer is 42.' } }
+  content_block_stop  { index: 1 }
+  message_delta       { delta: { stop_reason: 'end_turn' } }
+  message_stop
+```
+
+**工具调用**：
+```
+OpenAI chunks:
+  delta.tool_calls[0] = { id: 'call_xxx', function: { name: 'bash', arguments: '' } }
+  delta.tool_calls[0].function.arguments = '{"comm'
+  delta.tool_calls[0].function.arguments = 'and":"ls"}'
+  finish_reason = "tool_calls"
+
+→ Anthropic events:
+  message_start       { ... }
+  content_block_start { index: 0, content_block: { type: 'tool_use', id: 'call_xxx', name: 'bash' } }
+  content_block_delta { index: 0, delta: { type: 'input_json_delta', partial_json: '{"comm' } }
+  content_block_delta { index: 0, delta: { type: 'input_json_delta', partial_json: 'and":"ls"}' } }
+  content_block_stop  { index: 0 }
+  message_delta       { delta: { stop_reason: 'tool_use' } }
+  message_stop
+```
+
+## 功能支持
+
+### Thinking（思维链）
+
+**请求侧**：不需要显式配置。支持思维链的模型（DeepSeek 等）会自动返回 `delta.reasoning_content`。
+
+**响应侧**：`delta.reasoning_content` 被转换为 Anthropic `thinking` content block：
+
+```ts
+// content_block_start
+{ type: 'content_block_start', index: 0,
+  content_block: { type: 'thinking', thinking: '', signature: '' } }
+
+// content_block_delta
+{ type: 'content_block_delta', index: 0,
+  delta: { type: 'thinking_delta', thinking: 'Let me analyze...' } }
+```
+
+thinking block 在 text/tool_use block 之前自动关闭，保持 Anthropic 的块顺序。
+
+### Prompt Caching
+
+**请求侧**：OpenAI 端点使用自动缓存，无需显式设置 `cache_control`。
+
+**响应侧**：OpenAI 的 `usage.prompt_tokens_details.cached_tokens` 被映射到 Anthropic 的 `cache_read_input_tokens`：
+
+```
+OpenAI:   usage.prompt_tokens_details.cached_tokens = 800
+     ↓
+Anthropic: message_start.message.usage.cache_read_input_tokens = 800
+```
+
+在 `message_start` 的 usage 中报告缓存命中量。
+
+### 工具调用（Tool Use）
+
+完整支持 OpenAI function calling 格式。所有本地工具（Bash、FileEdit、Grep、Glob、Agent 等）透明工作——它们通过 JSON 输入输出通信，格式无关。
+
+工具参数以 `input_json_delta` 形式流式传输，由下游代码拼接解析。
+
+### 不支持的功能
+
+| 功能 | 策略 |
+|---|---|
+| Beta Headers | 不发送 |
+| Server Tools (advisor) | 不发送 |
+| Structured Output | 不发送 |
+| Fast Mode / Effort | 不发送 |
+| Tool Search / defer_loading | 不启用，所有工具直接发送 |
+| Anthropic Signature | thinking block 的 `signature` 字段为空字符串 |
+| cache_creation_input_tokens | 始终为 0（OpenAI 不区分创建/读取） |
+
+## 测试
+
+```bash
+# 运行所有 OpenAI 适配层测试
+bun test src/services/api/openai/__tests__/
+
+# 单独运行
+bun test src/services/api/openai/__tests__/streamAdapter.test.ts     # 14 tests（含 thinking + caching）
+bun test src/services/api/openai/__tests__/convertMessages.test.ts   # 10 tests
+bun test src/services/api/openai/__tests__/convertTools.test.ts      # 7 tests
+bun test src/services/api/openai/__tests__/modelMapping.test.ts      # 6 tests
+```
+
+当前测试覆盖：**39 tests / 73 assertions / 0 fail**。
+
+### 测试覆盖矩阵
+
+| 功能 | convertMessages | convertTools | streamAdapter | modelMapping |
+|---|---|---|---|---|
+| 文本消息转换 | ✅ | | | |
+| tool_use 转换 | ✅ | | | |
+| tool_result 转换 | ✅ | | | |
+| thinking 剥离 | ✅ | | | |
+| 完整对话流程 | ✅ | | | |
+| 工具 schema 转换 | | ✅ | | |
+| tool_choice 映射 | | ✅ | | |
+| 纯文本流 | | | ✅ | |
+| 工具调用流 | | | ✅ | |
+| 混合文本+工具 | | | ✅ | |
+| finish_reason 映射 | | | ✅ | |
+| thinking 流 | | | ✅ | |
+| thinking+text 切换 | | | ✅ | |
+| thinking+tool_use 切换 | | | ✅ | |
+| 块索引正确性 | | | ✅ | |
+| cached_tokens 映射 | | | ✅ | |
+| OPENAI_MODEL 覆盖 | | | | ✅ |
+| 默认模型映射 | | | | ✅ |
+| 未知模型透传 | | | | ✅ |
+| [1m] 后缀剥离 | | | | ✅ |
+
+## 端到端验证
+
+```bash
+# 1. 安装依赖
+bun install
+
+# 2. 运行单元测试
+bun test src/services/api/openai/__tests__/
+
+# 3. 连接实际端点（以 Ollama 为例）
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=ollama \
+OPENAI_BASE_URL=http://localhost:11434/v1 \
+OPENAI_MODEL=qwen2.5-coder-32b \
+bun run dev
+
+# 4. 连接 DeepSeek（测试 thinking 支持）
+CLAUDE_CODE_USE_OPENAI=1 \
+OPENAI_API_KEY=sk-xxx \
+OPENAI_BASE_URL=https://api.deepseek.com/v1 \
+OPENAI_MODEL=deepseek-reasoner \
+bun run dev
+
+# 5. 确认现有测试不受影响
+bun test  # 无 CLAUDE_CODE_USE_OPENAI 时走原有路径
+```
+
+## 代码统计
+
+| 类别 | 行数 |
+|---|---|
+| 新增源码 | ~620 行 |
+| 新增测试 | ~450 行 |
+| 改动现有代码 | ~25 行 |
+| **总计** | **~1100 行** |
--- a/package.json
+++ b/package.json
@@ -53,6 +53,7 @@
  },
  "dependencies": {},
  "devDependencies": {
+    "openai": "^4.73.0",
    "@alcalzone/ansi-tokenize": "^0.3.0",
    "@ant/claude-for-chrome-mcp": "workspace:*",
    "@ant/computer-use-input": "workspace:*",
--- a/src/components/ConsoleOAuthFlow.tsx
+++ b/src/components/ConsoleOAuthFlow.tsx
@@ -38,6 +38,15 @@ type OAuthStatus = {
  opusModel: string;
  activeField: 'base_url' | 'api_key' | 'haiku_model' | 'sonnet_model' | 'opus_model';
 } // Custom platform: configure API endpoint and model names
+| {
+  state: 'openai_chat_api';
+  baseUrl: string;
+  apiKey: string;
+  haikuModel: string;
+  sonnetModel: string;
+  opusModel: string;
+  activeField: 'base_url' | 'api_key' | 'haiku_model' | 'sonnet_model' | 'opus_model';
+} // OpenAI Chat Completions API platform
 | {
  state: 'ready_to_start';
 } // Flow started, waiting for browser to open
@@ -246,6 +255,8 @@ export function ConsoleOAuthFlow({
        if (!orgResult.valid) {
          throw new Error((orgResult as { valid: false; message: string }).message);
        }
+        // Reset modelType to anthropic when using OAuth login
+        updateSettingsForSource('userSettings', { modelType: 'anthropic' } as any);
        setOAuthStatus({
          state: 'success'
        });
@@ -416,6 +427,9 @@ function OAuthStatusMessage(t0) {
          t6 = [{
            label: <Text>Custom Platform ·{" "}<Text dimColor={true}>Configure your own API endpoint</Text>{"\n"}</Text>,
            value: "custom_platform"
+          }, {
+            label: <Text>OpenAI Compatible ·{" "}<Text dimColor={true}>Ollama, DeepSeek, vLLM, One API, etc.</Text>{"\n"}</Text>,
+            value: "openai_chat_api"
          }, t4, t5, {
            label: <Text>3rd-party platform ·{" "}<Text dimColor={true}>Amazon Bedrock, Microsoft Foundry, or Vertex AI</Text>{"\n"}</Text>,
            value: "platform"
@@ -438,6 +452,17 @@ function OAuthStatusMessage(t0) {
                  opusModel: process.env.ANTHROPIC_DEFAULT_OPUS_MODEL ?? "",
                  activeField: "base_url"
                });
+              } else if (value_0 === "openai_chat_api") {
+                logEvent("tengu_openai_chat_api_selected", {});
+                setOAuthStatus({
+                  state: "openai_chat_api",
+                  baseUrl: process.env.OPENAI_BASE_URL ?? "",
+                  apiKey: process.env.OPENAI_API_KEY ?? "",
+                  haikuModel: process.env.ANTHROPIC_DEFAULT_HAIKU_MODEL ?? "",
+                  sonnetModel: process.env.ANTHROPIC_DEFAULT_SONNET_MODEL ?? "",
+                  opusModel: process.env.ANTHROPIC_DEFAULT_OPUS_MODEL ?? "",
+                  activeField: "base_url"
+                });
              } else if (value_0 === "platform") {
                logEvent("tengu_oauth_platform_selected", {});
                setOAuthStatus({
@@ -568,7 +593,7 @@ function OAuthStatusMessage(t0) {
          if (finalVals.haiku_model) env.ANTHROPIC_DEFAULT_HAIKU_MODEL = finalVals.haiku_model;
          if (finalVals.sonnet_model) env.ANTHROPIC_DEFAULT_SONNET_MODEL = finalVals.sonnet_model;
          if (finalVals.opus_model) env.ANTHROPIC_DEFAULT_OPUS_MODEL = finalVals.opus_model;
-          const { error } = updateSettingsForSource('userSettings', { env } as any);
+          const { error } = updateSettingsForSource('userSettings', { modelType: 'anthropic' as any, env } as any);
          if (error) {
            setOAuthStatus({ state: 'error', message: `Failed to save: ${error.message}`, toRetry: { state: 'custom_platform', baseUrl: '', apiKey: '', haikuModel: '', sonnetModel: '', opusModel: '', activeField: 'base_url' } });
          } else {
@@ -639,6 +664,107 @@ function OAuthStatusMessage(t0) {
          <Text dimColor>Tab to switch · Enter on last field to save · Esc to go back</Text>
        </Box>;
      }
+    case "openai_chat_api":
+      {
+        type OpenAIField = 'base_url' | 'api_key' | 'haiku_model' | 'sonnet_model' | 'opus_model';
+        const OPENAI_FIELDS: OpenAIField[] = ['base_url', 'api_key', 'haiku_model', 'sonnet_model', 'opus_model'];
+        const op = oauthStatus as { state: 'openai_chat_api'; activeField: OpenAIField; baseUrl: string; apiKey: string; haikuModel: string; sonnetModel: string; opusModel: string };
+        const { activeField, baseUrl, apiKey, haikuModel, sonnetModel, opusModel } = op;
+        const openaiDisplayValues: Record<OpenAIField, string> = { base_url: baseUrl, api_key: apiKey, haiku_model: haikuModel, sonnet_model: sonnetModel, opus_model: opusModel };
+
+        const [openaiInputValue, setOpenaiInputValue] = useState(() => openaiDisplayValues[activeField]);
+        const [openaiInputCursorOffset, setOpenaiInputCursorOffset] = useState(() => openaiDisplayValues[activeField].length);
+
+        const buildOpenAIState = useCallback((field: OpenAIField, value: string, newActive?: OpenAIField) => {
+          const s = { state: 'openai_chat_api' as const, activeField: newActive ?? activeField, baseUrl, apiKey, haikuModel, sonnetModel, opusModel };
+          switch (field) {
+            case 'base_url': return { ...s, baseUrl: value };
+            case 'api_key': return { ...s, apiKey: value };
+            case 'haiku_model': return { ...s, haikuModel: value };
+            case 'sonnet_model': return { ...s, sonnetModel: value };
+            case 'opus_model': return { ...s, opusModel: value };
+          }
+        }, [activeField, baseUrl, apiKey, haikuModel, sonnetModel, opusModel]);
+
+        const doOpenAISave = useCallback(() => {
+          const finalVals = { ...openaiDisplayValues, [activeField]: openaiInputValue };
+          const env: Record<string, string> = {};
+          if (finalVals.base_url) env.OPENAI_BASE_URL = finalVals.base_url;
+          if (finalVals.api_key) env.OPENAI_API_KEY = finalVals.api_key;
+          if (finalVals.haiku_model) env.ANTHROPIC_DEFAULT_HAIKU_MODEL = finalVals.haiku_model;
+          if (finalVals.sonnet_model) env.ANTHROPIC_DEFAULT_SONNET_MODEL = finalVals.sonnet_model;
+          if (finalVals.opus_model) env.ANTHROPIC_DEFAULT_OPUS_MODEL = finalVals.opus_model;
+          const { error } = updateSettingsForSource('userSettings', { modelType: 'openai' as any, env } as any);
+          if (error) {
+            setOAuthStatus({ state: 'error', message: `Failed to save: ${error.message}`, toRetry: { state: 'openai_chat_api', baseUrl: '', apiKey: '', haikuModel: '', sonnetModel: '', opusModel: '', activeField: 'base_url' } });
+          } else {
+            for (const [k, v] of Object.entries(env)) process.env[k] = v;
+            setOAuthStatus({ state: 'success' });
+            void onDone();
+          }
+        }, [activeField, openaiInputValue, openaiDisplayValues, setOAuthStatus, onDone]);
+
+        const handleOpenAIEnter = useCallback(() => {
+          const idx = OPENAI_FIELDS.indexOf(activeField);
+          setOAuthStatus(buildOpenAIState(activeField, openaiInputValue));
+          if (idx === OPENAI_FIELDS.length - 1) {
+            doOpenAISave();
+          } else {
+            const next = OPENAI_FIELDS[idx + 1]!;
+            setOpenaiInputValue(openaiDisplayValues[next] ?? '');
+            setOpenaiInputCursorOffset((openaiDisplayValues[next] ?? '').length);
+          }
+        }, [activeField, openaiInputValue, buildOpenAIState, doOpenAISave, openaiDisplayValues, setOAuthStatus]);
+
+        useKeybinding('tabs:next', () => {
+          const idx = OPENAI_FIELDS.indexOf(activeField);
+          if (idx < OPENAI_FIELDS.length - 1) {
+            setOAuthStatus(buildOpenAIState(activeField, openaiInputValue, OPENAI_FIELDS[idx + 1]));
+            setOpenaiInputValue(openaiDisplayValues[OPENAI_FIELDS[idx + 1]!] ?? '');
+            setOpenaiInputCursorOffset((openaiDisplayValues[OPENAI_FIELDS[idx + 1]!] ?? '').length);
+          }
+        }, { context: 'Tabs' });
+        useKeybinding('tabs:previous', () => {
+          const idx = OPENAI_FIELDS.indexOf(activeField);
+          if (idx > 0) {
+            setOAuthStatus(buildOpenAIState(activeField, openaiInputValue, OPENAI_FIELDS[idx - 1]));
+            setOpenaiInputValue(openaiDisplayValues[OPENAI_FIELDS[idx - 1]!] ?? '');
+            setOpenaiInputCursorOffset((openaiDisplayValues[OPENAI_FIELDS[idx - 1]!] ?? '').length);
+          }
+        }, { context: 'Tabs' });
+        useKeybinding('confirm:no', () => {
+          setOAuthStatus({ state: 'idle' });
+        }, { context: 'Confirmation' });
+
+        const openaiColumns = useTerminalSize().columns - 20;
+
+        const renderOpenAIRow = (field: OpenAIField, label: string, opts?: { mask?: boolean }) => {
+          const active = activeField === field;
+          const val = openaiDisplayValues[field];
+          return <Box>
+            <Text backgroundColor={active ? 'suggestion' : undefined} color={active ? 'inverseText' : undefined}>{` ${label} `}</Text>
+            <Text> </Text>
+            {active
+              ? <TextInput value={openaiInputValue} onChange={setOpenaiInputValue} onSubmit={handleOpenAIEnter} cursorOffset={openaiInputCursorOffset} onChangeCursorOffset={setOpenaiInputCursorOffset} columns={openaiColumns} mask={opts?.mask ? "*" : undefined} focus={true} />
+              : (val
+                ? <Text color="success">{opts?.mask ? val.slice(0, 8) + '·'.repeat(Math.max(0, val.length - 8)) : val}</Text>
+                : null)}
+          </Box>;
+        };
+
+        return <Box flexDirection="column" gap={1}>
+          <Text bold={true}>OpenAI Compatible API Setup</Text>
+          <Text dimColor>Configure an OpenAI Chat Completions compatible endpoint (e.g. Ollama, DeepSeek, vLLM).</Text>
+          <Box flexDirection="column" gap={1}>
+            {renderOpenAIRow('base_url', 'Base URL ')}
+            {renderOpenAIRow('api_key', 'API Key  ', { mask: true })}
+            {renderOpenAIRow('haiku_model', 'Haiku    ')}
+            {renderOpenAIRow('sonnet_model', 'Sonnet   ')}
+            {renderOpenAIRow('opus_model', 'Opus     ')}
+          </Box>
+          <Text dimColor>Tab to switch · Enter on last field to save · Esc to go back</Text>
+        </Box>;
+      }
    case "waiting_for_login":
      {
        let t1;
--- a/src/services/api/claude.ts
+++ b/src/services/api/claude.ts
@@ -1301,6 +1301,15 @@ async function* queryModel(
    API_MAX_MEDIA_PER_REQUEST,
  )

+  // OpenAI-compatible provider: delegate to the OpenAI adapter layer
+  // after shared preprocessing (message normalization, tool filtering,
+  // media stripping) but before Anthropic-specific logic (betas, thinking, caching).
+  if (getAPIProvider() === 'openai') {
+    const { queryModelOpenAI } = await import('./openai/index.js')
+    yield* queryModelOpenAI(messagesForAPI, systemPrompt, filteredTools, signal, options)
+    return
+  }
+
  // Instrumentation: Track message count after normalization
  logEvent('tengu_api_after_normalize', {
    postNormalizedMessageCount: messagesForAPI.length,
--- a/src/services/api/openai/tests/convertMessages.test.ts
+++ b/src/services/api/openai/tests/convertMessages.test.ts
@@ -0,0 +1,157 @@
+import { describe, expect, test } from 'bun:test'
+import { anthropicMessagesToOpenAI } from '../convertMessages.js'
+import type { UserMessage, AssistantMessage } from '../../../../types/message.js'
+
+// Helpers to create internal-format messages
+function makeUserMsg(content: string | any[]): UserMessage {
+  return {
+    type: 'user',
+    uuid: '00000000-0000-0000-0000-000000000000',
+    message: { role: 'user', content },
+  } as UserMessage
+}
+
+function makeAssistantMsg(content: string | any[]): AssistantMessage {
+  return {
+    type: 'assistant',
+    uuid: '00000000-0000-0000-0000-000000000001',
+    message: { role: 'assistant', content },
+  } as AssistantMessage
+}
+
+describe('anthropicMessagesToOpenAI', () => {
+  test('converts system prompt to system message', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeUserMsg('hello')],
+      ['You are helpful.'] as any,
+    )
+    expect(result[0]).toEqual({ role: 'system', content: 'You are helpful.' })
+  })
+
+  test('joins multiple system prompt strings', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeUserMsg('hi')],
+      ['Part 1', 'Part 2'] as any,
+    )
+    expect(result[0]).toEqual({ role: 'system', content: 'Part 1\n\nPart 2' })
+  })
+
+  test('skips empty system prompt', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeUserMsg('hi')],
+      [] as any,
+    )
+    expect(result[0].role).toBe('user')
+  })
+
+  test('converts simple user text message', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeUserMsg('hello world')],
+      [] as any,
+    )
+    expect(result).toEqual([{ role: 'user', content: 'hello world' }])
+  })
+
+  test('converts user message with content array', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeUserMsg([
+        { type: 'text', text: 'line 1' },
+        { type: 'text', text: 'line 2' },
+      ])],
+      [] as any,
+    )
+    expect(result).toEqual([{ role: 'user', content: 'line 1\nline 2' }])
+  })
+
+  test('converts assistant message with text', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeAssistantMsg('response text')],
+      [] as any,
+    )
+    expect(result).toEqual([{ role: 'assistant', content: 'response text' }])
+  })
+
+  test('converts assistant message with tool_use', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeAssistantMsg([
+        { type: 'text', text: 'Let me help.' },
+        {
+          type: 'tool_use' as const,
+          id: 'toolu_123',
+          name: 'bash',
+          input: { command: 'ls' },
+        },
+      ])],
+      [] as any,
+    )
+    expect(result).toEqual([{
+      role: 'assistant',
+      content: 'Let me help.',
+      tool_calls: [{
+        id: 'toolu_123',
+        type: 'function',
+        function: { name: 'bash', arguments: '{"command":"ls"}' },
+      }],
+    }])
+  })
+
+  test('converts tool_result to tool message', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeUserMsg([
+        {
+          type: 'tool_result' as const,
+          tool_use_id: 'toolu_123',
+          content: 'file1.txt\nfile2.txt',
+        },
+      ])],
+      [] as any,
+    )
+    expect(result).toEqual([{
+      role: 'tool',
+      tool_call_id: 'toolu_123',
+      content: 'file1.txt\nfile2.txt',
+    }])
+  })
+
+  test('strips thinking blocks', () => {
+    const result = anthropicMessagesToOpenAI(
+      [makeAssistantMsg([
+        { type: 'thinking' as const, thinking: 'internal thoughts...' },
+        { type: 'text', text: 'visible response' },
+      ])],
+      [] as any,
+    )
+    expect(result).toEqual([{ role: 'assistant', content: 'visible response' }])
+  })
+
+  test('handles full conversation with tools', () => {
+    const result = anthropicMessagesToOpenAI(
+      [
+        makeUserMsg('list files'),
+        makeAssistantMsg([
+          {
+            type: 'tool_use' as const,
+            id: 'toolu_abc',
+            name: 'bash',
+            input: { command: 'ls' },
+          },
+        ]),
+        makeUserMsg([
+          {
+            type: 'tool_result' as const,
+            tool_use_id: 'toolu_abc',
+            content: 'file.txt',
+          },
+        ]),
+      ],
+      ['You are helpful.'] as any,
+    )
+
+    expect(result).toHaveLength(4)
+    expect(result[0].role).toBe('system')
+    expect(result[1].role).toBe('user')
+    expect(result[2].role).toBe('assistant')
+    expect((result[2] as any).tool_calls).toBeDefined()
+    expect(result[3].role).toBe('tool')
+  })
+})
--- a/src/services/api/openai/tests/convertTools.test.ts
+++ b/src/services/api/openai/tests/convertTools.test.ts
@@ -0,0 +1,85 @@
+import { describe, expect, test } from 'bun:test'
+import { anthropicToolsToOpenAI, anthropicToolChoiceToOpenAI } from '../convertTools.js'
+
+describe('anthropicToolsToOpenAI', () => {
+  test('converts basic tool', () => {
+    const tools = [
+      {
+        type: 'custom',
+        name: 'bash',
+        description: 'Run a bash command',
+        input_schema: {
+          type: 'object',
+          properties: { command: { type: 'string' } },
+          required: ['command'],
+        },
+      },
+    ]
+
+    const result = anthropicToolsToOpenAI(tools as any)
+
+    expect(result).toEqual([{
+      type: 'function',
+      function: {
+        name: 'bash',
+        description: 'Run a bash command',
+        parameters: {
+          type: 'object',
+          properties: { command: { type: 'string' } },
+          required: ['command'],
+        },
+      },
+    }])
+  })
+
+  test('uses empty schema when input_schema missing', () => {
+    const tools = [{ type: 'custom', name: 'noop', description: 'no-op' }]
+    const result = anthropicToolsToOpenAI(tools as any)
+
+    expect(result[0].function.parameters).toEqual({ type: 'object', properties: {} })
+  })
+
+  test('strips Anthropic-specific fields', () => {
+    const tools = [
+      {
+        type: 'custom',
+        name: 'bash',
+        description: 'Run bash',
+        input_schema: { type: 'object', properties: {} },
+        cache_control: { type: 'ephemeral' },
+        defer_loading: true,
+      },
+    ]
+    const result = anthropicToolsToOpenAI(tools as any)
+
+    expect((result[0] as any).cache_control).toBeUndefined()
+    expect((result[0] as any).defer_loading).toBeUndefined()
+  })
+
+  test('handles empty tools array', () => {
+    expect(anthropicToolsToOpenAI([])).toEqual([])
+  })
+})
+
+describe('anthropicToolChoiceToOpenAI', () => {
+  test('maps auto', () => {
+    expect(anthropicToolChoiceToOpenAI({ type: 'auto' })).toBe('auto')
+  })
+
+  test('maps any to required', () => {
+    expect(anthropicToolChoiceToOpenAI({ type: 'any' })).toBe('required')
+  })
+
+  test('maps tool to function', () => {
+    const result = anthropicToolChoiceToOpenAI({ type: 'tool', name: 'bash' })
+    expect(result).toEqual({ type: 'function', function: { name: 'bash' } })
+  })
+
+  test('returns undefined for undefined input', () => {
+    expect(anthropicToolChoiceToOpenAI(undefined)).toBeUndefined()
+  })
+
+  test('returns undefined for unknown type', () => {
+    expect(anthropicToolChoiceToOpenAI({ type: 'unknown' })).toBeUndefined()
+  })
+})
--- a/src/services/api/openai/tests/modelMapping.test.ts
+++ b/src/services/api/openai/tests/modelMapping.test.ts
@@ -0,0 +1,62 @@
+import { describe, expect, test, beforeEach, afterEach } from 'bun:test'
+import { resolveOpenAIModel } from '../modelMapping.js'
+
+describe('resolveOpenAIModel', () => {
+  const originalEnv = {
+    OPENAI_MODEL: process.env.OPENAI_MODEL,
+    ANTHROPIC_DEFAULT_HAIKU_MODEL: process.env.ANTHROPIC_DEFAULT_HAIKU_MODEL,
+    ANTHROPIC_DEFAULT_SONNET_MODEL: process.env.ANTHROPIC_DEFAULT_SONNET_MODEL,
+    ANTHROPIC_DEFAULT_OPUS_MODEL: process.env.ANTHROPIC_DEFAULT_OPUS_MODEL,
+  }
+
+  beforeEach(() => {
+    delete process.env.OPENAI_MODEL
+    delete process.env.ANTHROPIC_DEFAULT_HAIKU_MODEL
+    delete process.env.ANTHROPIC_DEFAULT_SONNET_MODEL
+    delete process.env.ANTHROPIC_DEFAULT_OPUS_MODEL
+  })
+
+  afterEach(() => {
+    Object.assign(process.env, originalEnv)
+  })
+
+  test('OPENAI_MODEL env var overrides all', () => {
+    process.env.OPENAI_MODEL = 'my-custom-model'
+    expect(resolveOpenAIModel('claude-sonnet-4-6')).toBe('my-custom-model')
+  })
+
+  test('ANTHROPIC_DEFAULT_SONNET_MODEL overrides default map', () => {
+    process.env.ANTHROPIC_DEFAULT_SONNET_MODEL = 'my-sonnet'
+    expect(resolveOpenAIModel('claude-sonnet-4-6')).toBe('my-sonnet')
+  })
+
+  test('ANTHROPIC_DEFAULT_HAIKU_MODEL overrides default map', () => {
+    process.env.ANTHROPIC_DEFAULT_HAIKU_MODEL = 'my-haiku'
+    expect(resolveOpenAIModel('claude-haiku-4-5-20251001')).toBe('my-haiku')
+  })
+
+  test('ANTHROPIC_DEFAULT_OPUS_MODEL overrides default map', () => {
+    process.env.ANTHROPIC_DEFAULT_OPUS_MODEL = 'my-opus'
+    expect(resolveOpenAIModel('claude-opus-4-6')).toBe('my-opus')
+  })
+
+  test('maps known Anthropic model via DEFAULT_MODEL_MAP', () => {
+    expect(resolveOpenAIModel('claude-sonnet-4-6')).toBe('gpt-4o')
+  })
+
+  test('maps haiku model', () => {
+    expect(resolveOpenAIModel('claude-haiku-4-5-20251001')).toBe('gpt-4o-mini')
+  })
+
+  test('maps opus model', () => {
+    expect(resolveOpenAIModel('claude-opus-4-6')).toBe('o3')
+  })
+
+  test('passes through unknown model name', () => {
+    expect(resolveOpenAIModel('some-random-model')).toBe('some-random-model')
+  })
+
+  test('strips [1m] suffix', () => {
+    expect(resolveOpenAIModel('claude-sonnet-4-6[1m]')).toBe('gpt-4o')
+  })
+})
--- a/src/services/api/openai/tests/streamAdapter.test.ts
+++ b/src/services/api/openai/tests/streamAdapter.test.ts
@@ -0,0 +1,434 @@
+import { describe, expect, test } from 'bun:test'
+import { adaptOpenAIStreamToAnthropic } from '../streamAdapter.js'
+import type { ChatCompletionChunk } from 'openai/resources/chat/completions/completions.mjs'
+
+/** Helper to create a mock async iterable from chunk array */
+function mockStream(chunks: ChatCompletionChunk[]): AsyncIterable<ChatCompletionChunk> {
+  return {
+    [Symbol.asyncIterator]() {
+      let i = 0
+      return {
+        async next() {
+          if (i >= chunks.length) return { done: true, value: undefined }
+          return { done: false, value: chunks[i++] }
+        },
+      }
+    },
+  }
+}
+
+/** Create a minimal ChatCompletionChunk */
+function makeChunk(overrides: Partial<ChatCompletionChunk> & any = {}): ChatCompletionChunk {
+  return {
+    id: 'chatcmpl-test',
+    object: 'chat.completion.chunk',
+    created: 1234567890,
+    model: 'gpt-4o',
+    choices: [],
+    ...overrides,
+  } as ChatCompletionChunk
+}
+
+async function collectEvents(chunks: ChatCompletionChunk[]) {
+  const events: any[] = []
+  for await (const event of adaptOpenAIStreamToAnthropic(mockStream(chunks), 'gpt-4o')) {
+    events.push(event)
+  }
+  return events
+}
+
+describe('adaptOpenAIStreamToAnthropic', () => {
+  test('emits message_start on first chunk', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { role: 'assistant', content: '' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { content: 'hello' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {},
+          finish_reason: 'stop',
+        }],
+        usage: { prompt_tokens: 10, completion_tokens: 5, total_tokens: 15 },
+      }),
+    ])
+
+    expect(events[0].type).toBe('message_start')
+    expect(events[0].message.role).toBe('assistant')
+    expect(events[0].message.model).toBe('gpt-4o')
+  })
+
+  test('converts text content stream', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{ index: 0, delta: { content: 'Hello' }, finish_reason: null }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: { content: ' world' }, finish_reason: null }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+      }),
+    ])
+
+    const types = events.map(e => e.type)
+    expect(types).toContain('message_start')
+    expect(types).toContain('content_block_start')
+    expect(types.filter(t => t === 'content_block_delta').length).toBe(2)
+    expect(types).toContain('content_block_stop')
+    expect(types).toContain('message_delta')
+    expect(types).toContain('message_stop')
+
+    const textDeltas = events.filter(e => e.type === 'content_block_delta') as any[]
+    expect(textDeltas[0].delta.text).toBe('Hello')
+    expect(textDeltas[1].delta.text).toBe(' world')
+  })
+
+  test('converts tool_calls stream', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {
+            tool_calls: [{
+              index: 0,
+              id: 'call_abc',
+              type: 'function',
+              function: { name: 'bash', arguments: '' },
+            }],
+          },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {
+            tool_calls: [{
+              index: 0,
+              function: { arguments: '{"comm' },
+            }],
+          },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {
+            tool_calls: [{
+              index: 0,
+              function: { arguments: 'and":"ls"}' },
+            }],
+          },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'tool_calls' }],
+      }),
+    ])
+
+    const blockStart = events.find(e => e.type === 'content_block_start') as any
+    expect(blockStart.content_block.type).toBe('tool_use')
+    expect(blockStart.content_block.name).toBe('bash')
+
+    const jsonDeltas = events.filter(
+      e => e.type === 'content_block_delta' && e.delta.type === 'input_json_delta',
+    ) as any[]
+    const fullArgs = jsonDeltas.map(d => d.delta.partial_json).join('')
+    expect(fullArgs).toBe('{"command":"ls"}')
+  })
+
+  test('maps finish_reason stop to end_turn', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{ index: 0, delta: { content: 'hi' }, finish_reason: null }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+      }),
+    ])
+
+    const msgDelta = events.find(e => e.type === 'message_delta') as any
+    expect(msgDelta.delta.stop_reason).toBe('end_turn')
+  })
+
+  test('maps finish_reason tool_calls to tool_use', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {
+            tool_calls: [{ index: 0, id: 'call_1', function: { name: 'bash', arguments: '{}' } }],
+          },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'tool_calls' }],
+      }),
+    ])
+
+    const msgDelta = events.find(e => e.type === 'message_delta') as any
+    expect(msgDelta.delta.stop_reason).toBe('tool_use')
+  })
+
+  test('maps finish_reason length to max_tokens', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{ index: 0, delta: { content: 'truncated' }, finish_reason: null }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'length' }],
+      }),
+    ])
+
+    const msgDelta = events.find(e => e.type === 'message_delta') as any
+    expect(msgDelta.delta.stop_reason).toBe('max_tokens')
+  })
+
+  test('handles mixed text and tool_calls', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{ index: 0, delta: { content: 'Thinking...' }, finish_reason: null }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {
+            tool_calls: [{ index: 0, id: 'call_1', function: { name: 'grep', arguments: '{"p":"test"}' } }],
+          },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'tool_calls' }],
+      }),
+    ])
+
+    const blockStarts = events.filter(e => e.type === 'content_block_start') as any[]
+    expect(blockStarts.length).toBe(2)
+    expect(blockStarts[0].content_block.type).toBe('text')
+    expect(blockStarts[1].content_block.type).toBe('tool_use')
+  })
+})
+
+describe('thinking support (reasoning_content)', () => {
+  test('converts reasoning_content to thinking block', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { reasoning_content: 'Let me analyze this...' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { reasoning_content: ' step by step.' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+      }),
+    ])
+
+    // Should have a thinking content block
+    const blockStart = events.find(e => e.type === 'content_block_start') as any
+    expect(blockStart.content_block.type).toBe('thinking')
+    expect(blockStart.content_block.signature).toBe('')
+
+    // Should have thinking_delta events
+    const thinkingDeltas = events.filter(
+      e => e.type === 'content_block_delta' && e.delta.type === 'thinking_delta',
+    ) as any[]
+    expect(thinkingDeltas.length).toBe(2)
+    expect(thinkingDeltas[0].delta.thinking).toBe('Let me analyze this...')
+    expect(thinkingDeltas[1].delta.thinking).toBe(' step by step.')
+  })
+
+  test('converts reasoning then content (DeepSeek-style)', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { reasoning_content: 'Thinking about the answer...' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { content: 'Here is my answer.' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+      }),
+    ])
+
+    // Should have two content blocks: thinking + text
+    const blockStarts = events.filter(e => e.type === 'content_block_start') as any[]
+    expect(blockStarts.length).toBe(2)
+    expect(blockStarts[0].content_block.type).toBe('thinking')
+    expect(blockStarts[1].content_block.type).toBe('text')
+
+    // Thinking block should be closed before text block starts
+    const blockStops = events.filter(e => e.type === 'content_block_stop') as any[]
+    expect(blockStops[0].index).toBe(0) // thinking block closed at index 0
+    expect(blockStarts[1].index).toBe(1) // text block starts at index 1
+
+    // Verify text delta
+    const textDelta = events.find(
+      e => e.type === 'content_block_delta' && e.delta.type === 'text_delta',
+    ) as any
+    expect(textDelta.delta.text).toBe('Here is my answer.')
+  })
+
+  test('handles reasoning then tool_calls', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { reasoning_content: 'I need to run a command.' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: {
+            tool_calls: [{ index: 0, id: 'call_1', function: { name: 'bash', arguments: '{"c":"ls"}' } }],
+          },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'tool_calls' }],
+      }),
+    ])
+
+    const blockStarts = events.filter(e => e.type === 'content_block_start') as any[]
+    expect(blockStarts.length).toBe(2)
+    expect(blockStarts[0].content_block.type).toBe('thinking')
+    expect(blockStarts[1].content_block.type).toBe('tool_use')
+  })
+
+  test('thinking block index is 0, text block index is 1', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { reasoning_content: 'reason' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { content: 'answer' },
+          finish_reason: null,
+        }],
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+      }),
+    ])
+
+    const blockStarts = events.filter(e => e.type === 'content_block_start') as any[]
+    expect(blockStarts[0].index).toBe(0)
+    expect(blockStarts[1].index).toBe(1)
+  })
+})
+
+describe('prompt caching support', () => {
+  test('maps cached_tokens to cache_read_input_tokens', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{
+          index: 0,
+          delta: { content: 'hi' },
+          finish_reason: null,
+        }],
+        usage: {
+          prompt_tokens: 1000,
+          completion_tokens: 0,
+          total_tokens: 1000,
+          prompt_tokens_details: { cached_tokens: 800 },
+        } as any,
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+        usage: {
+          prompt_tokens: 1000,
+          completion_tokens: 50,
+          total_tokens: 1050,
+          prompt_tokens_details: { cached_tokens: 800 },
+        } as any,
+      }),
+    ])
+
+    const msgStart = events.find(e => e.type === 'message_start') as any
+    expect(msgStart.message.usage.cache_read_input_tokens).toBe(800)
+    expect(msgStart.message.usage.input_tokens).toBe(1000)
+  })
+
+  test('defaults cache_read_input_tokens to 0 when no cached_tokens', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{ index: 0, delta: { content: 'hi' }, finish_reason: null }],
+        usage: { prompt_tokens: 100, completion_tokens: 0, total_tokens: 100 },
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+      }),
+    ])
+
+    const msgStart = events.find(e => e.type === 'message_start') as any
+    expect(msgStart.message.usage.cache_read_input_tokens).toBe(0)
+    expect(msgStart.message.usage.cache_creation_input_tokens).toBe(0)
+  })
+
+  test('updates cached_tokens from later chunks', async () => {
+    const events = await collectEvents([
+      makeChunk({
+        choices: [{ index: 0, delta: { content: 'hi' }, finish_reason: null }],
+        usage: {
+          prompt_tokens: 500,
+          completion_tokens: 0,
+          total_tokens: 500,
+        } as any,
+      }),
+      makeChunk({
+        choices: [{ index: 0, delta: {}, finish_reason: 'stop' }],
+        usage: {
+          prompt_tokens: 500,
+          completion_tokens: 10,
+          total_tokens: 510,
+          prompt_tokens_details: { cached_tokens: 300 },
+        } as any,
+      }),
+    ])
+
+    const msgStart = events.find(e => e.type === 'message_start') as any
+    // First chunk had no cached_tokens, so initially 0
+    // But the message_start usage reflects the first chunk's data
+    expect(msgStart.message.usage.cache_read_input_tokens).toBe(0)
+    expect(msgStart.message.usage.input_tokens).toBe(500)
+  })
+})
--- a/src/services/api/openai/client.ts
+++ b/src/services/api/openai/client.ts
@@ -0,0 +1,48 @@
+import OpenAI from 'openai'
+import { getProxyFetchOptions } from 'src/utils/proxy.js'
+import { isEnvTruthy } from '../../utils/envUtils.js'
+
+/**
+ * Environment variables:
+ *
+ * OPENAI_API_KEY: Required. API key for the OpenAI-compatible endpoint.
+ * OPENAI_BASE_URL: Recommended. Base URL for the endpoint (e.g. http://localhost:11434/v1).
+ * OPENAI_ORG_ID: Optional. Organization ID.
+ * OPENAI_PROJECT_ID: Optional. Project ID.
+ */
+
+let cachedClient: OpenAI | null = null
+
+export function getOpenAIClient(options?: {
+  maxRetries?: number
+  fetchOverride?: typeof fetch
+  source?: string
+}): OpenAI {
+  if (cachedClient) return cachedClient
+
+  const apiKey = process.env.OPENAI_API_KEY || ''
+  const baseURL = process.env.OPENAI_BASE_URL
+
+  const client = new OpenAI({
+    apiKey,
+    ...(baseURL && { baseURL }),
+    maxRetries: options?.maxRetries ?? 0,
+    timeout: parseInt(process.env.API_TIMEOUT_MS || String(600 * 1000), 10),
+    dangerouslyAllowBrowser: true,
+    ...(process.env.OPENAI_ORG_ID && { organization: process.env.OPENAI_ORG_ID }),
+    ...(process.env.OPENAI_PROJECT_ID && { project: process.env.OPENAI_PROJECT_ID }),
+    fetchOptions: getProxyFetchOptions({ forAnthropicAPI: false }) as RequestInit,
+    ...(options?.fetchOverride && { fetch: options.fetchOverride }),
+  })
+
+  if (!options?.fetchOverride) {
+    cachedClient = client
+  }
+
+  return client
+}
+
+/** Clear the cached client (useful when env vars change). */
+export function clearOpenAIClientCache(): void {
+  cachedClient = null
+}
--- a/src/services/api/openai/convertMessages.ts
+++ b/src/services/api/openai/convertMessages.ts
@@ -0,0 +1,184 @@
+import type {
+  BetaContentBlockParam,
+  BetaToolResultBlockParam,
+  BetaToolUseBlock,
+} from '@anthropic-ai/sdk/resources/beta/messages/messages.mjs'
+import type {
+  ChatCompletionAssistantMessageParam,
+  ChatCompletionMessageParam,
+  ChatCompletionSystemMessageParam,
+  ChatCompletionToolMessageParam,
+  ChatCompletionUserMessageParam,
+} from 'openai/resources/chat/completions/completions.mjs'
+import type { AssistantMessage, UserMessage } from '../../../types/message.js'
+import type { SystemPrompt } from '../../../utils/systemPromptType.js'
+
+/**
+ * Convert internal (UserMessage | AssistantMessage)[] to OpenAI-format messages.
+ *
+ * Key conversions:
+ * - system prompt → role: "system" message prepended
+ * - tool_use blocks → tool_calls[] on assistant message
+ * - tool_result blocks → role: "tool" messages
+ * - thinking blocks → silently dropped
+ * - cache_control → stripped
+ */
+export function anthropicMessagesToOpenAI(
+  messages: (UserMessage | AssistantMessage)[],
+  systemPrompt: SystemPrompt,
+): ChatCompletionMessageParam[] {
+  const result: ChatCompletionMessageParam[] = []
+
+  // Prepend system prompt as system message
+  const systemText = systemPromptToText(systemPrompt)
+  if (systemText) {
+    result.push({
+      role: 'system',
+      content: systemText,
+    } satisfies ChatCompletionSystemMessageParam)
+  }
+
+  for (const msg of messages) {
+    switch (msg.type) {
+      case 'user':
+        result.push(...convertInternalUserMessage(msg))
+        break
+      case 'assistant':
+        result.push(...convertInternalAssistantMessage(msg))
+        break
+      default:
+        break
+    }
+  }
+
+  return result
+}
+
+function systemPromptToText(systemPrompt: SystemPrompt): string {
+  if (!systemPrompt || systemPrompt.length === 0) return ''
+  return systemPrompt
+    .filter(Boolean)
+    .join('\n\n')
+}
+
+function convertInternalUserMessage(
+  msg: UserMessage,
+): ChatCompletionMessageParam[] {
+  const result: ChatCompletionMessageParam[] = []
+  const content = msg.message.content
+
+  if (typeof content === 'string') {
+    result.push({
+      role: 'user',
+      content,
+    } satisfies ChatCompletionUserMessageParam)
+  } else if (Array.isArray(content)) {
+    const textParts: string[] = []
+    const toolResults: BetaToolResultBlockParam[] = []
+
+    for (const block of content) {
+      if (typeof block === 'string') {
+        textParts.push(block)
+      } else if (block.type === 'text') {
+        textParts.push(block.text)
+      } else if (block.type === 'tool_result') {
+        toolResults.push(block as BetaToolResultBlockParam)
+      }
+      // Skip image, document, thinking, cache_edits, etc.
+    }
+
+    if (textParts.length > 0) {
+      result.push({
+        role: 'user',
+        content: textParts.join('\n'),
+      } satisfies ChatCompletionUserMessageParam)
+    }
+
+    for (const tr of toolResults) {
+      result.push(convertToolResult(tr))
+    }
+  }
+
+  return result
+}
+
+function convertToolResult(
+  block: BetaToolResultBlockParam,
+): ChatCompletionToolMessageParam {
+  let content: string
+  if (typeof block.content === 'string') {
+    content = block.content
+  } else if (Array.isArray(block.content)) {
+    content = block.content
+      .map(c => {
+        if (typeof c === 'string') return c
+        if ('text' in c) return c.text
+        return ''
+      })
+      .filter(Boolean)
+      .join('\n')
+  } else {
+    content = ''
+  }
+
+  return {
+    role: 'tool',
+    tool_call_id: block.tool_use_id,
+    content,
+  } satisfies ChatCompletionToolMessageParam
+}
+
+function convertInternalAssistantMessage(
+  msg: AssistantMessage,
+): ChatCompletionMessageParam[] {
+  const content = msg.message.content
+
+  if (typeof content === 'string') {
+    return [
+      {
+        role: 'assistant',
+        content,
+      } satisfies ChatCompletionAssistantMessageParam,
+    ]
+  }
+
+  if (!Array.isArray(content)) {
+    return [
+      {
+        role: 'assistant',
+        content: '',
+      } satisfies ChatCompletionAssistantMessageParam,
+    ]
+  }
+
+  const textParts: string[] = []
+  const toolCalls: NonNullable<ChatCompletionAssistantMessageParam['tool_calls']> = []
+
+  for (const block of content) {
+    if (typeof block === 'string') {
+      textParts.push(block)
+    } else if (block.type === 'text') {
+      textParts.push(block.text)
+    } else if (block.type === 'tool_use') {
+      const tu = block as BetaToolUseBlock
+      toolCalls.push({
+        id: tu.id,
+        type: 'function',
+        function: {
+          name: tu.name,
+          arguments:
+            typeof tu.input === 'string' ? tu.input : JSON.stringify(tu.input),
+        },
+      })
+    }
+    // Skip thinking, redacted_thinking, server_tool_use, etc.
+  }
+
+  const result: ChatCompletionAssistantMessageParam = {
+    role: 'assistant',
+    content: textParts.length > 0 ? textParts.join('\n') : null,
+    ...(toolCalls.length > 0 && { tool_calls: toolCalls }),
+  }
+
+  return [result]
+}
--- a/src/services/api/openai/convertTools.ts
+++ b/src/services/api/openai/convertTools.ts
@@ -0,0 +1,68 @@
+import type { BetaToolUnion } from '@anthropic-ai/sdk/resources/beta/messages/messages.mjs'
+import type { ChatCompletionTool } from 'openai/resources/chat/completions/completions.mjs'
+
+/**
+ * Convert Anthropic tool schemas to OpenAI function calling format.
+ *
+ * Anthropic: { name, description, input_schema }
+ * OpenAI:    { type: "function", function: { name, description, parameters } }
+ *
+ * Anthropic-specific fields (cache_control, defer_loading, etc.) are stripped.
+ */
+export function anthropicToolsToOpenAI(
+  tools: BetaToolUnion[],
+): ChatCompletionTool[] {
+  return tools
+    .filter(tool => {
+      // Only convert standard tools (skip server tools like computer_use, etc.)
+      return tool.type === 'custom' || !('type' in tool) || tool.type !== 'server'
+    })
+    .map(tool => {
+      // Handle the various tool shapes from Anthropic SDK
+      const anyTool = tool as Record<string, unknown>
+      const name = (anyTool.name as string) || ''
+      const description = (anyTool.description as string) || ''
+      const inputSchema = anyTool.input_schema as Record<string, unknown> | undefined
+
+      return {
+        type: 'function' as const,
+        function: {
+          name,
+          description,
+          parameters: inputSchema || { type: 'object', properties: {} },
+        },
+      } satisfies ChatCompletionTool
+    })
+}
+
+/**
+ * Map Anthropic tool_choice to OpenAI tool_choice format.
+ *
+ * Anthropic → OpenAI:
+ * - { type: "auto" } → "auto"
+ * - { type: "any" }  → "required"
+ * - { type: "tool", name } → { type: "function", function: { name } }
+ * - undefined → undefined (use provider default)
+ */
+export function anthropicToolChoiceToOpenAI(
+  toolChoice: unknown,
+): string | { type: 'function'; function: { name: string } } | undefined {
+  if (!toolChoice || typeof toolChoice !== 'object') return undefined
+
+  const tc = toolChoice as Record<string, unknown>
+  const type = tc.type as string
+
+  switch (type) {
+    case 'auto':
+      return 'auto'
+    case 'any':
+      return 'required'
+    case 'tool':
+      return {
+        type: 'function',
+        function: { name: tc.name as string },
+      }
+    default:
+      return undefined
+  }
+}
--- a/src/services/api/openai/index.ts
+++ b/src/services/api/openai/index.ts
@@ -0,0 +1,208 @@
+import type { BetaToolUnion } from '@anthropic-ai/sdk/resources/beta/messages/messages.mjs'
+import type { SystemPrompt } from '../../../utils/systemPromptType.js'
+import type { Message, StreamEvent, SystemAPIErrorMessage, AssistantMessage } from '../../../types/message.js'
+import type { Tools } from '../../../Tool.js'
+import { getOpenAIClient } from './client.js'
+import { anthropicMessagesToOpenAI } from './convertMessages.js'
+import { anthropicToolsToOpenAI, anthropicToolChoiceToOpenAI } from './convertTools.js'
+import { adaptOpenAIStreamToAnthropic } from './streamAdapter.js'
+import { resolveOpenAIModel } from './modelMapping.js'
+import { normalizeMessagesForAPI } from '../../../utils/messages.js'
+import { toolToAPISchema } from '../../../utils/api.js'
+import { getEmptyToolPermissionContext } from '../../../Tool.js'
+import { logForDebugging } from '../../../utils/debug.js'
+import type { Options } from '../claude.js'
+import { randomUUID } from 'crypto'
+import {
+  createAssistantAPIErrorMessage,
+  normalizeContentFromAPI,
+} from '../../../utils/messages.js'
+
+/**
+ * OpenAI-compatible query path. Converts Anthropic-format messages/tools to
+ * OpenAI format, calls the OpenAI-compatible endpoint, and converts the
+ * SSE stream back to Anthropic BetaRawMessageStreamEvent for consumption
+ * by the existing query pipeline.
+ */
+export async function* queryModelOpenAI(
+  messages: Message[],
+  systemPrompt: SystemPrompt,
+  tools: Tools,
+  signal: AbortSignal,
+  options: Options,
+): AsyncGenerator<
+  StreamEvent | AssistantMessage | SystemAPIErrorMessage,
+  void
+> {
+  try {
+    // 1. Resolve model name
+    const openaiModel = resolveOpenAIModel(options.model)
+
+    // 2. Normalize messages using shared preprocessing
+    const messagesForAPI = normalizeMessagesForAPI(messages, tools)
+
+    // 3. Build tool schemas
+    const toolSchemas = await Promise.all(
+      tools.map(tool =>
+        toolToAPISchema(tool, {
+          getToolPermissionContext: options.getToolPermissionContext,
+          tools,
+          agents: options.agents,
+          allowedAgentTypes: options.allowedAgentTypes,
+          model: options.model,
+        }),
+      ),
+    )
+    // Filter out non-standard tools (server tools like advisor)
+    const standardTools = toolSchemas.filter(
+      (t): t is BetaToolUnion & { type: string } => {
+        const anyT = t as Record<string, unknown>
+        return anyT.type !== 'advisor_20260301' && anyT.type !== 'computer_20250124'
+      },
+    )
+
+    // 4. Convert messages and tools to OpenAI format
+    const openaiMessages = anthropicMessagesToOpenAI(messagesForAPI, systemPrompt)
+    const openaiTools = anthropicToolsToOpenAI(standardTools)
+    const openaiToolChoice = anthropicToolChoiceToOpenAI(options.toolChoice)
+
+    // 5. Get client and make streaming request
+    const client = getOpenAIClient({
+      maxRetries: 0,
+      fetchOverride: options.fetchOverride,
+      source: options.querySource,
+    })
+
+    logForDebugging(`[OpenAI] Calling model=${openaiModel}, messages=${openaiMessages.length}, tools=${openaiTools.length}`)
+
+    // 6. Call OpenAI API with streaming
+    const stream = await client.chat.completions.create(
+      {
+        model: openaiModel,
+        messages: openaiMessages,
+        ...(openaiTools.length > 0 && {
+          tools: openaiTools,
+          ...(openaiToolChoice && { tool_choice: openaiToolChoice }),
+        }),
+        stream: true,
+        stream_options: { include_usage: true },
+        ...(options.temperatureOverride !== undefined && {
+          temperature: options.temperatureOverride,
+        }),
+      },
+      {
+        signal,
+      },
+    )
+
+    // 7. Convert OpenAI stream to Anthropic events, then process into
+    //    AssistantMessage + StreamEvent (matching the Anthropic path behavior)
+    const adaptedStream = adaptOpenAIStreamToAnthropic(stream, openaiModel)
+
+    // Accumulate content blocks and usage, same as the Anthropic path in claude.ts
+    const contentBlocks: Record<number, any> = {}
+    let partialMessage: any = undefined
+    let usage = {
+      input_tokens: 0,
+      output_tokens: 0,
+      cache_creation_input_tokens: 0,
+      cache_read_input_tokens: 0,
+    }
+    let ttftMs = 0
+    const start = Date.now()
+
+    for await (const event of adaptedStream) {
+      switch (event.type) {
+        case 'message_start': {
+          partialMessage = (event as any).message
+          ttftMs = Date.now() - start
+          if ((event as any).message?.usage) {
+            usage = {
+              ...usage,
+              ...((event as any).message.usage),
+            }
+          }
+          break
+        }
+        case 'content_block_start': {
+          const idx = (event as any).index
+          const cb = (event as any).content_block
+          if (cb.type === 'tool_use') {
+            contentBlocks[idx] = { ...cb, input: '' }
+          } else if (cb.type === 'text') {
+            contentBlocks[idx] = { ...cb, text: '' }
+          } else if (cb.type === 'thinking') {
+            contentBlocks[idx] = { ...cb, thinking: '', signature: '' }
+          } else {
+            contentBlocks[idx] = { ...cb }
+          }
+          break
+        }
+        case 'content_block_delta': {
+          const idx = (event as any).index
+          const delta = (event as any).delta
+          const block = contentBlocks[idx]
+          if (!block) break
+          if (delta.type === 'text_delta') {
+            block.text = (block.text || '') + delta.text
+          } else if (delta.type === 'input_json_delta') {
+            block.input = (block.input || '') + delta.partial_json
+          } else if (delta.type === 'thinking_delta') {
+            block.thinking = (block.thinking || '') + delta.thinking
+          } else if (delta.type === 'signature_delta') {
+            block.signature = delta.signature
+          }
+          break
+        }
+        case 'content_block_stop': {
+          const idx = (event as any).index
+          const block = contentBlocks[idx]
+          if (!block || !partialMessage) break
+
+          const m: AssistantMessage = {
+            message: {
+              ...partialMessage,
+              content: normalizeContentFromAPI(
+                [block],
+                tools,
+                options.agentId,
+              ),
+            },
+            requestId: undefined,
+            type: 'assistant',
+            uuid: randomUUID(),
+            timestamp: new Date().toISOString(),
+          }
+          yield m
+          break
+        }
+        case 'message_delta': {
+          const deltaUsage = (event as any).usage
+          if (deltaUsage) {
+            usage = { ...usage, ...deltaUsage }
+          }
+          // Update the stop_reason on the last yielded message
+          // (we don't have a reference here, but the consumer handles this)
+          break
+        }
+        case 'message_stop':
+          break
+      }
+
+      // Also yield as StreamEvent for real-time display (matching Anthropic path)
+      yield {
+        type: 'stream_event',
+        event,
+        ...(event.type === 'message_start' ? { ttftMs } : undefined),
+      } as StreamEvent
+    }
+  } catch (error) {
+    const errorMessage = error instanceof Error ? error.message : String(error)
+    logForDebugging(`[OpenAI] Error: ${errorMessage}`, { level: 'error' })
+    yield createAssistantAPIErrorMessage({
+      content: `API Error: ${errorMessage}`,
+      apiError: 'api_error',
+      error: error instanceof Error ? error : new Error(String(error)),
+    })
+  }
+}
--- a/src/services/api/openai/modelMapping.ts
+++ b/src/services/api/openai/modelMapping.ts
@@ -0,0 +1,56 @@
+/**
+ * Default mapping from Anthropic model names to OpenAI model names.
+ * Used only when ANTHROPIC_DEFAULT_*_MODEL env vars are not set.
+ */
+const DEFAULT_MODEL_MAP: Record<string, string> = {
+  'claude-sonnet-4-20250514': 'gpt-4o',
+  'claude-sonnet-4-5-20250929': 'gpt-4o',
+  'claude-sonnet-4-6': 'gpt-4o',
+  'claude-opus-4-20250514': 'o3',
+  'claude-opus-4-1-20250805': 'o3',
+  'claude-opus-4-5-20251101': 'o3',
+  'claude-opus-4-6': 'o3',
+  'claude-haiku-4-5-20251001': 'gpt-4o-mini',
+  'claude-3-5-haiku-20241022': 'gpt-4o-mini',
+  'claude-3-7-sonnet-20250219': 'gpt-4o',
+  'claude-3-5-sonnet-20241022': 'gpt-4o',
+}
+
+/**
+ * Determine the model family (haiku / sonnet / opus) from an Anthropic model ID.
+ */
+function getModelFamily(model: string): 'haiku' | 'sonnet' | 'opus' | null {
+  if (/haiku/i.test(model)) return 'haiku'
+  if (/opus/i.test(model)) return 'opus'
+  if (/sonnet/i.test(model)) return 'sonnet'
+  return null
+}
+
+/**
+ * Resolve the OpenAI model name for a given Anthropic model.
+ *
+ * Priority:
+ * 1. OPENAI_MODEL env var (override all)
+ * 2. ANTHROPIC_DEFAULT_{FAMILY}_MODEL env var (e.g. ANTHROPIC_DEFAULT_SONNET_MODEL)
+ * 3. DEFAULT_MODEL_MAP lookup
+ * 4. Pass through original model name
+ */
+export function resolveOpenAIModel(anthropicModel: string): string {
+  // Highest priority: explicit override
+  if (process.env.OPENAI_MODEL) {
+    return process.env.OPENAI_MODEL
+  }
+
+  // Strip [1m] suffix if present (Claude-specific modifier)
+  const cleanModel = anthropicModel.replace(/\[1m\]$/, '')
+
+  // Check ANTHROPIC_DEFAULT_*_MODEL env vars based on model family
+  const family = getModelFamily(cleanModel)
+  if (family) {
+    const envVar = `ANTHROPIC_DEFAULT_${family.toUpperCase()}_MODEL`
+    const override = process.env[envVar]
+    if (override) return override
+  }
+
+  return DEFAULT_MODEL_MAP[cleanModel] ?? cleanModel
+}
--- a/src/services/api/openai/streamAdapter.ts
+++ b/src/services/api/openai/streamAdapter.ts
@@ -0,0 +1,310 @@
+import type { BetaRawMessageStreamEvent } from '@anthropic-ai/sdk/resources/beta/messages/messages.mjs'
+import type { ChatCompletionChunk } from 'openai/resources/chat/completions/completions.mjs'
+import { randomUUID } from 'crypto'
+
+/**
+ * Adapt an OpenAI streaming response into Anthropic BetaRawMessageStreamEvent.
+ *
+ * Mapping:
+ *   First chunk              → message_start
+ *   delta.reasoning_content  → content_block_start(thinking) + thinking_delta + content_block_stop
+ *   delta.content            → content_block_start(text) + text_delta + content_block_stop
+ *   delta.tool_calls         → content_block_start(tool_use) + input_json_delta + content_block_stop
+ *   finish_reason            → message_delta(stop_reason) + message_stop
+ *   usage.cached_tokens      → cache_read_input_tokens in message_start usage
+ *
+ * Thinking support:
+ *   DeepSeek and compatible providers send `delta.reasoning_content` for chain-of-thought.
+ *   This is mapped to Anthropic's `thinking` content blocks:
+ *     content_block_start: { type: 'thinking', thinking: '', signature: '' }
+ *     content_block_delta: { type: 'thinking_delta', thinking: '...' }
+ *
+ * Prompt caching:
+ *   OpenAI reports cached tokens in usage.prompt_tokens_details.cached_tokens.
+ *   This is mapped to Anthropic's cache_read_input_tokens.
+ */
+export async function* adaptOpenAIStreamToAnthropic(
+  stream: AsyncIterable<ChatCompletionChunk>,
+  model: string,
+): AsyncGenerator<BetaRawMessageStreamEvent, void> {
+  const messageId = `msg_${randomUUID().replace(/-/g, '').slice(0, 24)}`
+
+  let started = false
+  let currentContentIndex = -1
+
+  // Track tool_use blocks: tool_calls index → { contentIndex, id, name, arguments }
+  const toolBlocks = new Map<number, { contentIndex: number; id: string; name: string; arguments: string }>()
+
+  // Track thinking block state
+  let thinkingBlockOpen = false
+
+  // Track text block state
+  let textBlockOpen = false
+
+  // Track usage
+  let inputTokens = 0
+  let outputTokens = 0
+  let cachedTokens = 0
+
+  // Track all open content block indices (for cleanup)
+  const openBlockIndices = new Set<number>()
+
+  for await (const chunk of stream) {
+    const choice = chunk.choices?.[0]
+    const delta = choice?.delta
+
+    // Extract usage from any chunk that carries it
+    if (chunk.usage) {
+      inputTokens = chunk.usage.prompt_tokens ?? inputTokens
+      outputTokens = chunk.usage.completion_tokens ?? outputTokens
+      // OpenAI prompt caching: prompt_tokens_details.cached_tokens
+      const details = (chunk.usage as any).prompt_tokens_details
+      if (details?.cached_tokens) {
+        cachedTokens = details.cached_tokens
+      }
+    }
+
+    // Emit message_start on first chunk
+    if (!started) {
+      started = true
+
+      yield {
+        type: 'message_start',
+        message: {
+          id: messageId,
+          type: 'message',
+          role: 'assistant',
+          content: [],
+          model,
+          stop_reason: null,
+          stop_sequence: null,
+          usage: {
+            input_tokens: inputTokens,
+            output_tokens: 0,
+            cache_creation_input_tokens: 0,
+            cache_read_input_tokens: cachedTokens,
+          },
+        },
+      } as BetaRawMessageStreamEvent
+    }
+
+    if (!delta) continue
+
+    // Handle reasoning_content → Anthropic thinking block
+    // DeepSeek and compatible providers send delta.reasoning_content
+    const reasoningContent = (delta as any).reasoning_content
+    if (reasoningContent != null && reasoningContent !== '') {
+      if (!thinkingBlockOpen) {
+        currentContentIndex++
+        thinkingBlockOpen = true
+        openBlockIndices.add(currentContentIndex)
+
+        yield {
+          type: 'content_block_start',
+          index: currentContentIndex,
+          content_block: {
+            type: 'thinking',
+            thinking: '',
+            signature: '',
+          },
+        } as BetaRawMessageStreamEvent
+      }
+
+      yield {
+        type: 'content_block_delta',
+        index: currentContentIndex,
+        delta: {
+          type: 'thinking_delta',
+          thinking: reasoningContent,
+        },
+      } as BetaRawMessageStreamEvent
+    }
+
+    // Handle text content
+    if (delta.content != null && delta.content !== '') {
+      if (!textBlockOpen) {
+        // Close thinking block if still open (reasoning done, now generating answer)
+        if (thinkingBlockOpen) {
+          yield {
+            type: 'content_block_stop',
+            index: currentContentIndex,
+          } as BetaRawMessageStreamEvent
+          openBlockIndices.delete(currentContentIndex)
+          thinkingBlockOpen = false
+        }
+
+        currentContentIndex++
+        textBlockOpen = true
+        openBlockIndices.add(currentContentIndex)
+
+        yield {
+          type: 'content_block_start',
+          index: currentContentIndex,
+          content_block: {
+            type: 'text',
+            text: '',
+          },
+        } as BetaRawMessageStreamEvent
+      }
+
+      yield {
+        type: 'content_block_delta',
+        index: currentContentIndex,
+        delta: {
+          type: 'text_delta',
+          text: delta.content,
+        },
+      } as BetaRawMessageStreamEvent
+    }
+
+    // Handle tool calls
+    if (delta.tool_calls) {
+      for (const tc of delta.tool_calls) {
+        const tcIndex = tc.index
+
+        if (!toolBlocks.has(tcIndex)) {
+          // Close thinking block if open
+          if (thinkingBlockOpen) {
+            yield {
+              type: 'content_block_stop',
+              index: currentContentIndex,
+            } as BetaRawMessageStreamEvent
+            openBlockIndices.delete(currentContentIndex)
+            thinkingBlockOpen = false
+          }
+
+          // Close text block if open
+          if (textBlockOpen) {
+            yield {
+              type: 'content_block_stop',
+              index: currentContentIndex,
+            } as BetaRawMessageStreamEvent
+            openBlockIndices.delete(currentContentIndex)
+            textBlockOpen = false
+          }
+
+          // Start new tool_use block
+          currentContentIndex++
+          const toolId = tc.id || `toolu_${randomUUID().replace(/-/g, '').slice(0, 24)}`
+          const toolName = tc.function?.name || ''
+
+          toolBlocks.set(tcIndex, {
+            contentIndex: currentContentIndex,
+            id: toolId,
+            name: toolName,
+            arguments: '',
+          })
+          openBlockIndices.add(currentContentIndex)
+
+          yield {
+            type: 'content_block_start',
+            index: currentContentIndex,
+            content_block: {
+              type: 'tool_use',
+              id: toolId,
+              name: toolName,
+              input: {},
+            },
+          } as BetaRawMessageStreamEvent
+        }
+
+        // Stream argument fragments
+        const argFragment = tc.function?.arguments
+        if (argFragment) {
+          toolBlocks.get(tcIndex)!.arguments += argFragment
+          yield {
+            type: 'content_block_delta',
+            index: toolBlocks.get(tcIndex)!.contentIndex,
+            delta: {
+              type: 'input_json_delta',
+              partial_json: argFragment,
+            },
+          } as BetaRawMessageStreamEvent
+        }
+      }
+    }
+
+    // Handle finish
+    if (choice?.finish_reason) {
+      // Close thinking block if still open
+      if (thinkingBlockOpen) {
+        yield {
+          type: 'content_block_stop',
+          index: currentContentIndex,
+        } as BetaRawMessageStreamEvent
+        openBlockIndices.delete(currentContentIndex)
+        thinkingBlockOpen = false
+      }
+
+      // Close text block if still open
+      if (textBlockOpen) {
+        yield {
+          type: 'content_block_stop',
+          index: currentContentIndex,
+        } as BetaRawMessageStreamEvent
+        openBlockIndices.delete(currentContentIndex)
+        textBlockOpen = false
+      }
+
+      // Close all tool blocks that haven't been closed yet
+      for (const [, block] of toolBlocks) {
+        if (openBlockIndices.has(block.contentIndex)) {
+          yield {
+            type: 'content_block_stop',
+            index: block.contentIndex,
+          } as BetaRawMessageStreamEvent
+          openBlockIndices.delete(block.contentIndex)
+        }
+      }
+
+      // Map finish_reason to Anthropic stop_reason
+      const stopReason = mapFinishReason(choice.finish_reason)
+
+      yield {
+        type: 'message_delta',
+        delta: {
+          stop_reason: stopReason,
+          stop_sequence: null,
+        },
+        usage: {
+          output_tokens: outputTokens,
+        },
+      } as BetaRawMessageStreamEvent
+
+      yield {
+        type: 'message_stop',
+      } as BetaRawMessageStreamEvent
+    }
+  }
+
+  // Safety: close any remaining open blocks if stream ended without finish_reason
+  for (const idx of openBlockIndices) {
+    yield {
+      type: 'content_block_stop',
+      index: idx,
+    } as BetaRawMessageStreamEvent
+  }
+}
+
+/**
+ * Map OpenAI finish_reason to Anthropic stop_reason.
+ *
+ * stop           → end_turn
+ * tool_calls     → tool_use
+ * length         → max_tokens
+ * content_filter → end_turn
+ */
+function mapFinishReason(reason: string): string {
+  switch (reason) {
+    case 'stop':
+      return 'end_turn'
+    case 'tool_calls':
+      return 'tool_use'
+    case 'length':
+      return 'max_tokens'
+    case 'content_filter':
+      return 'end_turn'
+    default:
+      return 'end_turn'
+  }
+}
--- a/src/utils/model/configs.ts
+++ b/src/utils/model/configs.ts
@@ -11,6 +11,7 @@ export const CLAUDE_3_7_SONNET_CONFIG = {
  bedrock: 'us.anthropic.claude-3-7-sonnet-20250219-v1:0',
  vertex: 'claude-3-7-sonnet@20250219',
  foundry: 'claude-3-7-sonnet',
+  openai: 'claude-3-7-sonnet-20250219',
 } as const satisfies ModelConfig

 export const CLAUDE_3_5_V2_SONNET_CONFIG = {
@@ -18,6 +19,7 @@ export const CLAUDE_3_5_V2_SONNET_CONFIG = {
  bedrock: 'anthropic.claude-3-5-sonnet-20241022-v2:0',
  vertex: 'claude-3-5-sonnet-v2@20241022',
  foundry: 'claude-3-5-sonnet',
+  openai: 'claude-3-5-sonnet-20241022',
 } as const satisfies ModelConfig

 export const CLAUDE_3_5_HAIKU_CONFIG = {
@@ -25,6 +27,7 @@ export const CLAUDE_3_5_HAIKU_CONFIG = {
  bedrock: 'us.anthropic.claude-3-5-haiku-20241022-v1:0',
  vertex: 'claude-3-5-haiku@20241022',
  foundry: 'claude-3-5-haiku',
+  openai: 'claude-3-5-haiku-20241022',
 } as const satisfies ModelConfig

 export const CLAUDE_HAIKU_4_5_CONFIG = {
@@ -32,6 +35,7 @@ export const CLAUDE_HAIKU_4_5_CONFIG = {
  bedrock: 'us.anthropic.claude-haiku-4-5-20251001-v1:0',
  vertex: 'claude-haiku-4-5@20251001',
  foundry: 'claude-haiku-4-5',
+  openai: 'claude-haiku-4-5-20251001',
 } as const satisfies ModelConfig

 export const CLAUDE_SONNET_4_CONFIG = {
@@ -39,6 +43,7 @@ export const CLAUDE_SONNET_4_CONFIG = {
  bedrock: 'us.anthropic.claude-sonnet-4-20250514-v1:0',
  vertex: 'claude-sonnet-4@20250514',
  foundry: 'claude-sonnet-4',
+  openai: 'claude-sonnet-4-20250514',
 } as const satisfies ModelConfig

 export const CLAUDE_SONNET_4_5_CONFIG = {
@@ -46,6 +51,7 @@ export const CLAUDE_SONNET_4_5_CONFIG = {
  bedrock: 'us.anthropic.claude-sonnet-4-5-20250929-v1:0',
  vertex: 'claude-sonnet-4-5@20250929',
  foundry: 'claude-sonnet-4-5',
+  openai: 'claude-sonnet-4-5-20250929',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_CONFIG = {
@@ -53,6 +59,7 @@ export const CLAUDE_OPUS_4_CONFIG = {
  bedrock: 'us.anthropic.claude-opus-4-20250514-v1:0',
  vertex: 'claude-opus-4@20250514',
  foundry: 'claude-opus-4',
+  openai: 'claude-opus-4-20250514',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_1_CONFIG = {
@@ -60,6 +67,7 @@ export const CLAUDE_OPUS_4_1_CONFIG = {
  bedrock: 'us.anthropic.claude-opus-4-1-20250805-v1:0',
  vertex: 'claude-opus-4-1@20250805',
  foundry: 'claude-opus-4-1',
+  openai: 'claude-opus-4-1-20250805',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_5_CONFIG = {
@@ -67,6 +75,7 @@ export const CLAUDE_OPUS_4_5_CONFIG = {
  bedrock: 'us.anthropic.claude-opus-4-5-20251101-v1:0',
  vertex: 'claude-opus-4-5@20251101',
  foundry: 'claude-opus-4-5',
+  openai: 'claude-opus-4-5-20251101',
 } as const satisfies ModelConfig

 export const CLAUDE_OPUS_4_6_CONFIG = {
@@ -74,6 +83,7 @@ export const CLAUDE_OPUS_4_6_CONFIG = {
  bedrock: 'us.anthropic.claude-opus-4-6-v1',
  vertex: 'claude-opus-4-6',
  foundry: 'claude-opus-4-6',
+  openai: 'claude-opus-4-6',
 } as const satisfies ModelConfig

 export const CLAUDE_SONNET_4_6_CONFIG = {
@@ -81,6 +91,7 @@ export const CLAUDE_SONNET_4_6_CONFIG = {
  bedrock: 'us.anthropic.claude-sonnet-4-6',
  vertex: 'claude-sonnet-4-6',
  foundry: 'claude-sonnet-4-6',
+  openai: 'claude-sonnet-4-6',
 } as const satisfies ModelConfig

 // @[MODEL LAUNCH]: Register the new config here.
--- a/src/utils/model/providers.ts
+++ b/src/utils/model/providers.ts
@@ -1,16 +1,24 @@
 import type { AnalyticsMetadata_I_VERIFIED_THIS_IS_NOT_CODE_OR_FILEPATHS } from '../../services/analytics/index.js'
+import { getInitialSettings } from '../settings/settings.js'
 import { isEnvTruthy } from '../envUtils.js'

-export type APIProvider = 'firstParty' | 'bedrock' | 'vertex' | 'foundry'
+export type APIProvider = 'firstParty' | 'bedrock' | 'vertex' | 'foundry' | 'openai'

 export function getAPIProvider(): APIProvider {
-  return isEnvTruthy(process.env.CLAUDE_CODE_USE_BEDROCK)
-    ? 'bedrock'
-    : isEnvTruthy(process.env.CLAUDE_CODE_USE_VERTEX)
-      ? 'vertex'
-      : isEnvTruthy(process.env.CLAUDE_CODE_USE_FOUNDRY)
-        ? 'foundry'
-        : 'firstParty'
+  // 1. Check settings.json modelType field (highest priority)
+  const modelType = getInitialSettings().modelType
+  if (modelType === 'openai') return 'openai'
+
+  // 2. Check environment variables (backward compatibility)
+  return isEnvTruthy(process.env.CLAUDE_CODE_USE_OPENAI)
+    ? 'openai'
+    : isEnvTruthy(process.env.CLAUDE_CODE_USE_BEDROCK)
+      ? 'bedrock'
+      : isEnvTruthy(process.env.CLAUDE_CODE_USE_VERTEX)
+        ? 'vertex'
+        : isEnvTruthy(process.env.CLAUDE_CODE_USE_FOUNDRY)
+          ? 'foundry'
+          : 'firstParty'
 }

 export function getAPIProviderForStatsig(): AnalyticsMetadata_I_VERIFIED_THIS_IS_NOT_CODE_OR_FILEPATHS {
--- a/src/utils/settings/types.ts
+++ b/src/utils/settings/types.ts
@@ -372,6 +372,13 @@ export const SettingsSchema = lazySchema(() =>
      permissions: PermissionsSchema()
        .optional()
        .describe('Tool usage permissions configuration'),
+      modelType: z
+        .enum(['anthropic', 'openai'])
+        .optional()
+        .describe(
+          'API provider type. "anthropic" uses the Anthropic API (default), "openai" uses the OpenAI Chat Completions API (/v1/chat/completions). ' +
+            'When set to "openai", configure OPENAI_API_KEY, OPENAI_BASE_URL, and OPENAI_MODEL in env.',
+        ),
      model: z
        .string()
        .optional()
--- a/tests/integration/cli-arguments.test.ts
+++ b/tests/integration/cli-arguments.test.ts
@@ -1,11 +1,10 @@
 import { describe, expect, test } from "bun:test";
+import { Command } from "@commander-js/extra-typings";

 // Test Commander.js option parsing independently from main.tsx initialization.
 // main.tsx has heavy bootstrap dependencies; we test the CLI argument parsing
 // patterns it uses to ensure correct behavior.

-const { Command } = require("/Users/konghayao/code/ai/claude-code/node_modules/.old_modules-13e6b62a502cda34/commander/index.js");
-
 function createTestProgram(): Command {
  const program = new Command();
  program