ollama/ollama

Known for

Getting a model running locally without reading a setup guide. One command, then an API on localhost.

Local model running reduced to a single command. The on-ramp for most self-hosted AI, and the reason a large share of people ever run a model on their own hardware.