AI repository intelligence

FR-AI / ANALYSIS

Why it is worth attention

It bridges proprietary M365 Copilot with open API standards, enabling developers to use powerful Copilot models via familiar OpenAI or Anthropic interfaces, and includes advanced features like MCP protocol and multi-account management.

Who it is for

Developers integrating M365 Copilot into existing AI workflows
Administrators managing multiple Copilot accounts and API keys
Users who want to access Copilot models through standard chat completion APIs
Builders of custom AI applications needing gateway functionality

Use cases

Use as a drop-in backend for any OpenAI-compatible client (e.g., ChatGPT apps)
Build chatbots and tools that leverage M365 Copilot reasoning models
Integrate Copilot into MCP-aware applications for dynamic tool execution
Create a controlled API layer with account pooling and proxy rotation

Strengths

OpenAI and Anthropic API compatibility with streaming and tool calling
Supports multiple Copilot models including reasoning variants (gpt-5.5-reasoning, gpt-5.6-sol)
Includes MCP protocol endpoints for model context protocol
Provides web admin console, API key management, and proxy pool for production use

Considerations

Requires valid M365 Copilot subscription – not a standalone AI service
Models and features are limited to those exposed by the Copilot backend
Security depends on proper configuration (default binds to localhost, admin password change recommended)

README quick start

M365 Copilot2API

M365 Copilot ChatHub 网关将 Microsoft 365 Copilot 转换为 OpenAI / Anthropic 兼容 API

📋 功能特性

功能	状态
✅ OpenAI 兼容 `/v1/chat/completions`	支持流式、工具调用
✅ OpenAI Responses `/v1/responses`	兼容
✅ Anthropic 兼容 `/v1/messages`	兼容
✅ 流式输出 (SSE)	实时逐字输出
✅ 推理模型	gpt-5.5-reasoning、gpt-5.6-sol 等
✅ 联网搜索	claude-sonnet 内置 web_fetch
✅ MCP 协议	`/v1/mcp/sse` + `/v1/mcp/message` + `/v1/mcp/tools`
✅ 视觉识别	支持 base64 图片输入
✅ 多账号管理	PKCE OAuth 授权
✅ API Key 管理	控制台创建/撤销
✅ 代理池	轮换、失败冷却
✅ Web 管理控制台	账号、设置、日志

🚀 快速开始

源码编译

git clone https://github.com/HEXUXIU/M365-Copilot2API.git
cd M365-Copilot2API
# 配置管理员密码（可选，默认 admin123）
export M365_ADMIN_PASSWORD=your_password
go run ./cmd/server

默认地址 http://127.0.0.1:4141，打开浏览器完成管理员设置和登录。

Docker 部署

docker compose build
docker compose up -d

🔑 使用示例

# 基础聊天
curl http://127.0.0.1:4141/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.5",
    "messages": [{"role": "user", "content": "你好"}]
  }'

# 流式输出
curl http://127.0.0.1:4141/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.5-reasoning",
    "messages": [{"role": "user", "content": "1+1=?"}],
    "stream": true
  }'

# 联网搜索（claude-sonnet 内置）
curl http://127.0.0.1:4141/v1/chat/completions \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "claude-sonnet",
    "messages": [{"role": "user", "content": "北京今天天气？"}]
  }'

🤖 可用模型

模型	推荐用途	速度
`gpt-5.5`	日常对话	⚡ ~5s
`gpt-5.2`	最轻量快速	⚡ ~5s
`claude-sonnet`	联网搜索、实时信息	⚡ ~5-6s
`gpt-5.6-sol`	复杂推理（默认高推理）	⏳ ~7s
`gpt-5.5-reasoning`	数学/逻辑推理	⏳ ~9s

🧩 MCP 协议

网关内置 MCP (Model Context Protocol) 服务器：

端点	说明
`GET /v1/mcp/sse`	SSE 连接
`POST /v1/mcp/message`	JSON-RPC 消息
`GET /v1/mcp/tools`	工具列表
`tools/list`	发现工具
`tools/call`	调用工具

🛠 技术栈

语言: Go 1.22+
协议: SignalR / WebSocket / SSE / MCP
前端: 单页 HTML + Inter 字体 + Lucide 图标
部署: Docker / 裸机

📄 配置

变量	默认值	说明
`M365_LISTEN`	`

Description

Access M365 Copilot via CLI & OpenAI-compatible API

Related repositories

Similar projects matched by category, topics, and programming language.

lopopolo

Featured

harness-engineering

Harness Engineering is a methodology for improving coding agent outputs by carefully crafting the environment around them—providing curated context, tools, and executable constraints that encode an organization’s nonfunctional requirements and cumulative lessons.

AI & Machine LearningAI Agents

2,390

slvDev

Featured

esp32-ai

A 28.9 million parameter language model runs on an $8 ESP32-S3 microcontroller entirely on-device, generating simple stories at about 9.5 tokens per second.

AI & Machine LearningLarge Language Models

1,960

littledivy

Featured

mimic

mimic captures traffic from any iOS or web app and automatically generates a Python client library that lets you call the app's API like a regular library.

AI & Machine Learning

1,482