Copilot Proxy

A local proxy server for GitHub Copilot that provides API compatibility with OpenAI's API format.

Installation

macOS/Linux (curl)

curl -sSL instl.vercel.app/copilot-proxy | bash

Windows

powershell -c "irm https://instl.vercel.app/copilot-proxy?platform=windows|iex"

You can also use docker image see compose file in repo.

Features

• OpenAI API-compatible endpoints for chat completions, embeddings, and models
• GitHub Copilot authentication via device flow
• CORS support for web applications
• In-memory token caching
• Streaming support for chat completions
• Vision API support

Quick Start

1. Authenticate with GitHub Copilot:

copilot-proxy auth

This will provide you with an authentication token.

2. Start the proxy server:

copilot-proxy server

By default, the server runs on port 8080.

Configuration

Machine ID

The proxy requires a VSCode Machine ID to function. This ID is automatically stored in:

• Windows: C:\Users\<username>\AppData\Roaming\Code\User\globalStorage\storage.json
• macOS: ~/Library/Application Support/Code/User/globalStorage/storage.json
• Linux: ~/.config/Code/User/globalStorage/storage.json

To use your existing VSCode Machine ID:

1. Locate the storage.json file for your platform
2. Find the telemetry.machineId value in the file
3. Set it as the VSCODE_MACHINE_ID environment variable

Environment variables:

• PORT: Server port (default: 8080)
• COPILOT_VERSION: Copilot version (default: 1.277.0)
• COPILOT_API_VERSION: API version (default: 2025年01月21日)
• COPILOT_EDITOR_VERSION: Editor version (default: JetBrains-IC/243.25659.59)
• COPILOT_EDITOR_PLUGIN_VERSION: Plugin version (default: copilot-intellij/1.5.38-243)
• COPILOT_INTEGRATION_ID: Integration ID (default: jetbrains-chat)
• COPILOT_CLIENT_ID: Client ID for GitHub OAuth (default: Iv23ctfURkiMfJ4xr5mv)

Mimic Vscode environment

To mimic the VSCode environment, set the following environment variables:

• COPILOT_EDITOR_VERSION: vscode/1.99.2
• COPILOT_EDITOR_PLUGIN_VERSION: copilot-chat/0.27.2025041102
• COPILOT_INTEGRATION_ID: vscode-chat
• COPILOT_CLIENT_ID: Iv1.b507a08c87ecfe98
• COPILOT_VERSION: 0.27.2025041102

API Usage

Authentication

Include the token from copilot auth in your requests:

curl -H "Authorization: Bearer your-token" http://localhost:8080/v1/models

Available Endpoints

POST /v1/chat/completions

Compatible with OpenAI's chat completions API. Supports:

• Regular chat completions
• Streaming responses
• Vision API (include images in messages)

GET /v1/models

Lists available models.

POST /v1/embeddings

Generate embeddings for text.

Examples

Chat Completion

curl -X POST http://localhost:8080/v1/chat/completions \
 -H "Authorization: Bearer your-token" \
 -H "Content-Type: application/json" \
 -d '{
 "model": "gpt-4",
 "messages": [{"role": "user", "content": "Hello!"}]
 }'

Vision API

curl -X POST http://localhost:8080/v1/chat/completions \
 -H "Authorization: Bearer your-token" \
 -H "Content-Type: application/json" \
 -d '{
 "model": "gpt-4o",
 "messages": [{
 "role": "user",
 "content": [
 {
 "type": "text",
 "text": "What's in this image?"
 },
 {
 "type": "image_url",
 "image_url": "https://example.com/image.jpg"
 }
 ]
 }]
 }'

Building from Source

git clone https://codeberg.org/elevatedai/copilot-proxy.git
cd copilot-proxy
go build ./cmd/copilot-proxy

License

MIT License