Image AI

Local Image Classification

MobileNet v2 runs entirely in your browser. The model loads once (~14 MB from Hugging Face CDN), then all inference is local. Your images are never uploaded.

First load downloads the model from Hugging Face CDN. Subsequent uses may serve from browser cache. No image data is ever transmitted to a server.

Image Classification

Upload an image to run local AI inference using ViT Base (~86 MB). The model loads once and runs entirely in your browser.

Click to upload or drag an image here

Image stays in your browser. Not uploaded to any server.