Copilot API Proxy

Warning

This project proxies GitHub Copilot into OpenAI-compatible and Anthropic-compatible endpoints. It is unofficial and may break if upstream behavior changes.

Overview

This service exposes GitHub Copilot through a small compatibility layer so it can be used by tools expecting OpenAI or Anthropic style APIs.

Compared with the original upstream project, this fork keeps the README intentionally simpler and includes support for the responses passthrough endpoint.

Features

• OpenAI-compatible endpoints for chat, models, embeddings, and responses
• Anthropic-compatible messages endpoint
• Optional downstream API key protection from CLI arg or environment variable
• Usage and token inspection endpoints
• Optional rate limit control and manual approval flow
• Support for individual, business, and enterprise Copilot accounts

Installation

bun install

Run

Development:

bun run dev

Production:

bun run start

Common Commands

• Build: bun run build
• Lint: bun run lint
• Test: bun test
• Start: bun run start

API Key Protection

You can require clients to send an API key for all incoming requests.

Priority order:

• CLI arg --api-key
• env API_KEY
• env COPILOT_API_KEY
• default empty, meaning disabled

Example:

bun run start -- --port 3000 --api-key my-secret-key

Then call the API with either header:

• Authorization: Bearer my-secret-key
• x-api-key: my-secret-key

API Endpoints

OpenAI-compatible

Endpoint	Method	Notes
/chat/completions	POST	Chat completions passthrough
/v1/chat/completions	POST	Chat completions passthrough
/embeddings	POST	Embeddings passthrough
/v1/embeddings	POST	Embeddings passthrough
/models	GET	Model list
/v1/models	GET	Model list
/responses	POST	Responses passthrough
/v1/responses	POST	Responses passthrough

Anthropic-compatible

Endpoint	Method	Notes
/v1/messages	POST	Anthropic messages compatibility
/v1/messages/count_tokens	POST	Token counting

Utility

Endpoint	Method	Notes
/usage	GET	Usage information
/token	GET	Current Copilot token

responses Support

This fork adds direct passthrough for the OpenAI-style responses API:

• POST /responses
• POST /v1/responses

The server forwards the incoming request body to Copilot's responses endpoint and returns the upstream response body, headers, and status code directly.

Notes

• Requires Bun and a valid GitHub Copilot subscription
• Authentication and runtime behavior still follow the existing project implementation
• Use responsibly and avoid abusive automated traffic patterns