LocalAI

A small web app that lets you chat with AI models running entirely on your own machine — no internet, no third-party API. It talks to locally downloaded models through Ollama’s API endpoint, wrapped in a simple Flask + Jinja interface. Built to understand how local inference works end to end.