WebGPU LLM In The Browser

N0X runs open-source chat models directly in supported browsers through WebGPU and MLC WebLLM. Models download once, cache locally, and stream responses without an account or hosted inference backend.