PilotPilot is your self hosted, privacy first coding assistant for Visual Studio Code. It brings AI powered chat and inline code completions right into your editor helping you code, debug, and explore your project faster. Best of all, your code stays yours no data ever leaves your machine. Features
RequirementsTo use Pilot, you must run a local or remote Ollama server with 2 types of models, a normal LLM for coding and an Embedding LLM for indexing. The ones I recommend are:
You can also use other models. Install them with:
Then, set the server URL and model names in your VS Code settings. ⚠️ Heads up: If your machine isn’t powerful enough to run these models smoothly, you may notice delays during chat, completion, or indexing Extension SettingsYou can configure Pilot through your VS Code settings (Cmd/Ctrl + ,, then search for "Pilot"):
Known Issues
Release Notes1.0.0
ContributingExcited to have contributors on board! While contribution guidelines are still in progress, feel free to try out the extension and report any issues. LicenseMIT © Nart Aliti Pilot is all about giving you control over your AI assistant—right inside your favorite editor. Happy coding! 🧑💻 |