ollama-copilot

Why we included this project

Ollama Copilot is a small Go proxy that gives you GitHub Copilot-style inline completions while keeping the model choice in your hands. It translates the Copilot API your editor expects into requests any Ollama instance can serve, so a local codellama:code model works out of the box, and hosted providers like DeepSeek or Mistral are available when you want stronger suggestions. The setup script builds the binary, installs it under ~/.local/bin, and offers to configure Neovim, VSCode, and Zed for you, with an optional systemd service for running it in the background. For teams that already run Ollama and would rather not hand their code to a third-party service, this is a low-friction way to get copilot-style completions with few moving parts.