An OpenAI API compatible server that uses GitHub Copilot SDK for LLMs
An OpenAI API compatible server that uses GitHub Copilot SDK for LLMs
I've been playing around with the new official GitHub Copilot SDK and realized it's a goldmine for building programmatic bridges to their models. I built this server in Go to act as a OpenAI-compatible proxy. It essentially lets you treat your GitHub Copilot subscription as a standard OpenAI backend for any tool that supports it. I have tested it against OpenWebUI and Langchain. Key Highlights: - Official SDK: Built using the new Github Copilot SDK. It’s much more robust than the reverse-engineered solutions floating around and does not use unpublished APIs. - Tool Calling Support: It maps OpenAI function definitions to Copilot's agentic tools. You can use your own tools/functions through the Copilot without copilot needing access to the said tools just the definitions is enough. The goal was to create a reliable "bridge" so I can use my subscription models in my preferred interfaces.
Share cardActual performance
Launch Intel predictions
Analyze your own launch →Incorrect prediction on native model
Similar products
GitHub Copilot in JupyterLab
GitHub Copilot => OpenAI API Proxy. Serverless
Custom LLMs, without training. Use via openai compatible api
OpenAI-Compatible API at 1/15th the Price
Free GitHub Copilot CLI with your own model or API
OfoxAI — One API key for 100+ LLMs. OpenAI-compatible
GitHub Copilot for SQL
Voice in, voice out — with GitHub Copilot
Level Up Coding – Flying with GitHub Copilot
GitHub Copilot Is Helping Me Learn RxJS