Addi — Extend Copilot with Your Own AI Models

Addi bridges GitHub Copilot with your preferred AI providers. Instead of being locked into a single vendor's models, you can connect Copilot to OpenAI-compatible APIs, Anthropic Claude, Google Gemini, or any custom LLM endpoint — giving you full control over your AI stack while keeping Copilot's interface and workflow.

Features

• Multi-Provider Support — Connect OpenAI, Anthropic, Google, DeepSeek, Ollama, or any OpenAI-compatible API
• Full Copilot Integration — Models appear directly in Copilot's model picker
• Streaming Responses — Real-time token-by-token output
• Tool Calling — Use Copilot tools (code search, terminal, file editing) with your models
• Reasoning / Thinking — View the model's reasoning process for supported models
• Vision — Send images to models that support multimodal input
• Speed Monitoring — Track real-time performance metrics per model
• Import / Export — Backup, migrate, or share configurations as JSON

Quick Start

1. Install the extension from the VS Code Marketplace
2. Open the Addi sidebar in the Activity Bar
3. Add a Provider — click the + button, enter your API endpoint and key
4. Add a Model — right-click the provider, fill in the model ID and capabilities
5. Open Copilot Chat — select your model from the model picker and start chatting

For detailed setup instructions, see the User Guide.

Supported Providers

Documentation

Contributing

Issues and pull requests are welcome at github.com/deepwn/addi.

License

MIT License