LFM2.5-VL-1.6B WebGPU

Vision-Language Model in Your Browser

This demo showcases in-browser vision-language inference with LFM2.5-VL-1.6B, powered by ONNX Runtime and WebGPU.

Everything runs entirely in your browser with WebGPU acceleration, meaning no data is sent to a server.

Liquid

LFM2.5-VL-1.6B Stream from your webcam with real-time captions,powered by your own hardware with WebGPU!

Input:

Idle

Select quantization:

0 MB

Start capturing to see live captions...