dispatchAI/browser-chat
0
dispatchAI Browser Chat — WebGPU On-Device Inference
A 100% browser-based AI chat running SmolLM2-360M entirely on the user's GPU via WebGPU.
- Zero backend — static HTML/JS, infinitely scalable, costs nothing ever
- Zero data leaves the device — pure client-side inference
- WebGPU accelerated — falls back to WASM on unsupported browsers
- ~200MB model — downloads once, cached by the browser
This is the purest expression of dispatchAI's mission: AI that runs on your device, no cloud required.