🖼️ Image Classify

Upload a photo and a Vision Transformer classifies it — entirely in your browser. No image ever leaves your device. Model: Xenova/vit-base-patch16-224 (Google ViT-Base, ONNX) via Transformers.js.

📁 Drop an image here, or click to choose

JPG / PNG / WebP — processed locally

Try:

Loading model… (first load downloads ~330 MB, cached after)

How it works: the page loads @xenova/transformers from a CDN and the ViT model weights from the Hugging Face Hub. Inference runs with ONNX Runtime Web (WebAssembly) on your CPU. This is a Static Space — there is no server and no GPU involved, which is why it works on a free account.