🚀 LiteLLM Connector for GitHub Copilot Chat

Bring any LiteLLM-supported model into the Copilot Chat model picker — OpenAI, Anthropic (Claude), Google, Mistral, local Llama, and more.

If LiteLLM can talk to it, Copilot can use it.

⭐️ Support the project

If you find this useful, please:

• Star on GitHub: https://github.com/gethnet/litellm-connector-copilot
• Leave a rating/review on the VS Code Marketplace: https://marketplace.visualstudio.com/items?itemName=GethNet.litellm-connector-copilot
• Rate on Open VSX: https://open-vsx.org/extension/GethNet/litellm-connector-copilot

Want to support development?

• Ko-fi: https://ko-fi.com/amwdrizz
• Buy Me a Coffee: https://buymeacoffee.com/amwdrizz

🚨 Troubleshooting: Connection Issues 🚨

If the extension fails to connect or models don't show up:

1. Manual Setup: Run LiteLLM: Manage Configuration from the Command Palette (Ctrl+Shift+P). This often fixes setup "hiccups".
2. Verify: Run LiteLLM: Check Connection to test your settings.
3. Reset: If things are totally stuck, run LiteLLM: Reset All Configuration. This is the "nuke" option to clear all state.
4. Note: Reinstalling usually won't help as settings are stored securely in VS Code. Use the Reset command instead.

✅ Requirements

• �️ VS Code 1.120+ (required)
• �🔑 GitHub Copilot subscription (Free plan works)
• 🌐 A LiteLLM proxy URL and API key

⚡ Quick Start (60 seconds)

1. Install GitHub Copilot Chat
2. Install LiteLLM Connector for Copilot
3. Open Command Palette (Cmd/Ctrl+Shift+P)
4. Run LiteLLM: Manage Configuration
5. Add one or more backends with:
   • Name (used as the provider/group label)
   • Base URL (example: http://localhost:4000)
   • API Key (required)
6. Open Copilot Chat → pick a model under your configured backend/provider group → chat

Preferred configuration path: Use the VS Code Language Models provider-group flow (modern config) for full VS Code 1.120+ model-picker/category behavior. Legacy settings remain compatibility fallbacks.

Multi-Backend: Configure multiple LiteLLM provider groups to aggregate models from several proxies. Models are namespaced and grouped by backend so provider identity is clear.

Backend identity remains clear: each backend is kept as its own provider/group so model origin is obvious in the picker.

⚠️ 2.0.0 temporary limitation: Unless manually configured, Git commit message generation and inline completions are currently inoperative in this release.
Manual setup: use LiteLLM: Select Commit Message Model and LiteLLM: Select Inline Completion Model if you need these features now.

✨ What you get

• 🌍 Hundreds of models via your LiteLLM proxy
• ⛓️ Multi-backend aggregation from multiple proxy instances
• 🌊 Real-time streaming responses
• 🛠️ Tool / function calling support
• 👁️ Vision models supported (where available)
• 🧠 Unified Chat Provider — supports thinking parts for reasoning models via VS Code 1.120 APIs
• 🧠 Smart parameter handling for model quirks
• 🔁 Automatic retry when a model rejects unsupported flags
• 📊 Token tracking & usage monitoring for input/output tokens
• ✍️ Git commit generation from staged changes in the SCM view (requires manual configuration in 2.0.0)
• 🧼 Smart Sanitization automatically strips Markdown code blocks from generated commit messages
• 🔍 Connection diagnostics to verify proxy configuration
• ⏱️ Inactivity watchdog to prevent stuck streams
• 🚫🧠 Cache bypass controls (no-cache headers) with provider-aware behavior
• 🔐 Secure credential storage using VS Code SecretStorage
• ⌨️ Optional inline completions via VS Code's stable inline completion API (requires manual configuration in 2.0.0)

🆕 Recent Highlights

• 📝 Multi-Repo Commit Generation (correctly identifies the active repository in multi-repo workspaces)
• 🧪 Telemetry & Observability (PostHog-backed feature tracking and structured JSONL logging)
• 🔧 Model Capability Overrides (manually override toolCalling and imageInput detection)
• 🧠 Unified Chat Provider (supports thinking parts and newer VS Code 1.120 chat APIs)
• 🧼 SCM Message Sanitization (automatically cleans up generated commit messages by stripping triple backticks)
• ✍️ Git Commit Message Generation (generate messages from staged changes directly in the SCM view)
• 📊 Enhanced Token Awareness (real-time token counting and context window display in model tooltips)
• 🔍 Connection Diagnostics (new Check Connection command to validate proxy settings)
• 🚀 VS Code 1.120+ settings modernization (aligns with Language Model provider-group configuration)
• 🧱 Tool-call compatibility hardening (normalizes tool call IDs to OpenAI-compatible limits)
• 🧰 Stability Improvements (hardened JSON parsing and stream error recovery)
• 📦 Smaller, faster package (bundled/minified production builds)

⚙️ Configuration

⌨️ Commands

• LiteLLM: Manage Configuration: Opens LiteLLM backend management (multi-backend flow).
• LiteLLM: Check Connection: Verify proxy URL and API key configuration.
• LiteLLM: Select Inline Completion Model: Choose a model for inline completions.
• LiteLLM: Select Commit Message Model: Choose a model for git commit generation.
• LiteLLM: Show Available Models: See all models currently discovered from your proxy.
• LiteLLM: Reload Models: Manually refresh the model list.
• LiteLLM: Reset All Configuration: Clear all stored URLs and API keys.

🐛 Bug reports & feature requests

Please use GitHub Issues: https://github.com/gethnet/litellm-connector-copilot/issues

Including VS Code version, extension version, model id, and LiteLLM proxy details/logs (if possible) helps reproduce issues quickly.

🧩 Notes

• This extension is a provider for the official Copilot Chat experience.
• It won't function without the GitHub Copilot Chat extension installed.

🆘 Support

• Issues & feedback: https://github.com/gethnet/litellm-connector-copilot/issues
• License: Apache-2.0