Load a model above to start chatting
All inference runs locally in your browser — no data is sent to any server.