GitHub Copilot OpenAI-Compatible Provider for AI SDK

This package provides an OpenAI-compatible interface for GitHub Copilot API, designed to work seamlessly with Vercel AI SDK v6.

Features

• Full TypeScript support
• Seamless integration with Vercel AI SDK v6
• Easy to use API matching other AI SDK providers
• Automatic endpoint switching: Uses /responses endpoint for Codex models, /chat/completions for others
• Smart request formatting: Automatically converts messages array to OpenAI Responses API input format for Codex models

Installation

AI SDK v6 (Current)

npm install @opeoginni/github-copilot-openai-compatible

AI SDK v5

If you need AI SDK v5 support, use the ai-v5 tag:

npm install @opeoginni/github-copilot-openai-compatible@ai-v5

Authentication

Getting Your Token

To get your GitHub Copilot API token, check out opencode-copilot-auth.

Required Headers

GitHub Copilot requires specific headers for authentication. While the provider handles your API key, you may need to configure additional headers:

const githubCopilot = createGithubCopilotOpenAICompatible({
  apiKey: process.env.COPILOT_TOKEN,
  headers: {
    "Copilot-Integration-Id": "vscode-chat",
    "User-Agent": "GitHubCopilotChat/0.26.7",
    "Editor-Version": "vscode/1.104.1",
    "Editor-Plugin-Version": "copilot-chat/0.26.7"
  },
});

These headers identify your application to GitHub Copilot. You may need to update version numbers based on your integration.

Usage

Quick Start

import { createGithubCopilotOpenAICompatible } from '@opeoginni/github-copilot-openai-compatible';
import { generateText } from 'ai';

const githubCopilot = createGithubCopilotOpenAICompatible({
  apiKey: process.env.COPILOT_TOKEN,
  headers: {
    "Copilot-Integration-Id": "vscode-chat",
    "User-Agent": "GitHubCopilotChat/0.26.7",
    "Editor-Version": "vscode/1.104.1",
    "Editor-Plugin-Version": "copilot-chat/0.26.7"
  },
});

// Use Codex model (recommended)
const { text } = await generateText({
  model: githubCopilot('gpt-5.1-codex'),
  prompt: 'Write a Python function to sort a list',
});

console.log(text);

Supported Models

Codex Models (Uses /responses endpoint)

This package fully supports OpenAI's Codex models, which use the advanced OpenAI Responses API:

• gpt-5-codex - GPT-5 Codex
• gpt-5.1-codex - GPT-5.1 Codex
• gpt-5.1-codex-mini - GPT-5.1 Codex Mini
• gpt-5.1-codex-max - GPT-5.1 Codex Max

Other Models

For non-Codex models (standard chat completions), check your GitHub Copilot settings to see which models are available to you. You can use any model ID that Copilot supports - they'll automatically route to the /chat/completions endpoint.

Note: GitHub Copilot may provide access to various models (Claude, GPT, Gemini, etc.) based on your subscription. Check your Copilot settings for the full list of available models.

How It Works

Endpoint Routing

The provider automatically routes requests to the correct endpoint based on the model ID:

• Codex Models (gpt-5-codex or any model containing 'codex'):

  • Uses /responses endpoint
  • Converts messages array to OpenAI Responses API input format
  • Transforms message content parts (e.g., text → input_text, image_url → input_image)
  • System messages become developer role messages

• All Other Models:

  • Uses standard /chat/completions endpoint
  • Uses standard OpenAI-compatible messages array format

This means you don't need to worry about the underlying API differences - the provider handles it automatically!

API Reference

createGithubCopilotOpenAICompatible(options)

Creates a new GitHub Copilot provider instance.

Options:

• apiKey?: Your GitHub Copilot API token
• baseURL?: Base URL for API calls (default: https://api.githubcopilot.com)
• name?: Provider name (default: githubcopilot)
• headers?: Custom headers to include in requests
• fetch?: Custom fetch implementation

Returns: A provider instance.

githubCopilot

Pre-configured default provider instance. For production use, create your own instance with proper headers:

import { githubCopilot } from '@opeoginni/github-copilot-openai-compatible';
import { generateText } from 'ai';

// Create your own configured instance
const copilot = createGithubCopilotOpenAICompatible({
  apiKey: process.env.COPILOT_TOKEN,
  headers: {
    "Copilot-Integration-Id": "vscode-chat",
    "User-Agent": "GitHubCopilotChat/0.26.7",
    "Editor-Version": "vscode/1.104.1",
    "Editor-Plugin-Version": "copilot-chat/0.26.7"
  },
});

const { text } = await generateText({
  model: copilot('gpt-5.1-codex'),
  prompt: 'Hello, world!',
});

License

MIT