SmolVLM

Vision-language model running entirely in your browser. No server required.

Model HuggingFaceTB/SmolVLM-256M-Instruct

Parameters 256M

Quantization q4 (~150 MB download)

Runtime WebGPU (falls back to WASM)

Framework Transformers.js

Click "Load Model" to begin.

Drop image here or click to upload

PNG, JPG, WebP, GIF

Output will appear here.