Ampersand Bridge for GitHub Copilot Chat

Use ai& models directly from the GitHub Copilot Chat model picker in Visual Studio Code.

This extension is a native VS Code LanguageModelChatProvider. It validates your ai& API key, discovers the hosted models available to it, and streams OpenAI-compatible chat completions directly from https://api.aiand.com/v1 into Copilot Chat.

Ampersand Bridge is community-maintained and independent: it is not affiliated with, endorsed by, or supported by ai&. For the officially maintained experience, consider ai& for GitHub Copilot (aiand.aiand-copilot). See the security model for the exact network surface.

Highlights

• Direct ai& integration without a local proxy
• API keys owned by VS Code native provider configuration
• Multiple isolated entries with stable explicit IDs in Manage Language Models
• Live /v1/models discovery with durable per-credential catalog snapshots
• Streaming text, reasoning output, token usage, and function-tool calls
• Per-model reasoning-effort lists that never send an unsupported effort
• Bounded retries for pre-stream transient failures only
• ai& credit balance and usage tracking with a status-bar item and quick-pick details
• Best-effort per-token pricing in the model picker, sourced from live API data or bundled estimates
• Opt-in fill-in-the-middle inline suggestions with measured model badges and reasoning suppressed

Quick start

1. Install the extension. You need VS Code 1.125 or newer and GitHub Copilot Chat.
2. Create an API key in the ai& console, or run Ampersand Bridge: Open API Keys.
3. Open Copilot Chat, select Manage Models, add an Ampersand Bridge provider entry, and enter a unique lowercase Entry ID plus the key.
4. Choose any model returned for that key.

To use more than one key, add another Ampersand Bridge entry with a different Entry ID. Keep that ID when rotating the key; saved model selections stay stable and old request handles are retired. Use Ampersand Bridge: Manage Connection to select entries independently for management and inline suggestions, and to refresh models, test inference, or show usage. Before deleting an entry, run Ampersand Bridge: Forget Native Entry and remove it in Manage Language Models.

Documentation

• Setup, settings, and troubleshooting
• Models and pricing
• API key and security model
• Development and releases

Related projects

• Codex Bridge for Copilot Chat — Use OpenAI Codex models in Copilot Chat with a ChatGPT Plus or Pro subscription.
• Grok for GitHub Copilot Chat — Use xAI Grok models directly from the GitHub Copilot Chat model picker.
• Inception for GitHub Copilot Chat — Use Inception Mercury models directly from the GitHub Copilot Chat model picker.
• Ollama Cloud for GitHub Copilot Chat — Use Ollama Cloud models with native thinking and tool support.
• OpenCode for GitHub Copilot Chat — Use OpenCode Zen, Go, and Console models from the model picker.
• Orvix for GitHub Copilot Chat — Use Orvix managed and BYOK models directly from the GitHub Copilot Chat model picker.
• Poolside for GitHub Copilot Chat — Use hosted Poolside coding models directly from the GitHub Copilot Chat model picker.

Unofficial project; not affiliated with ai&, GitHub, or Microsoft. ai& usage limits and charges still apply. Licensed under MIT.