What AI model is used?

This tool uses CLIP ViT-Base-Patch32, published by OpenAI and served via Hugging Face as Xenova/clip-vit-base-patch32. CLIP (Contrastive Language–Image Pre-Training) matches images against text labels using a shared embedding space, making it ideal for zero-shot image classification. It runs entirely in your browser via @huggingface/transformers (transformers.js) — your images are never sent to a server.

Why is the first load slow?

The CLIP ViT-Base-Patch32 model is approximately 150 MB. On your first visit it downloads from Hugging Face and is cached permanently in your browser's storage. All subsequent uses load instantly from cache with no internet connection needed for the model.

Is my image uploaded anywhere?

No. Everything runs locally in your browser using WebAssembly (ONNX Runtime Web). Your image never leaves your device — it is processed entirely on your own machine.

AI Image Tagger — Auto Tag Photos Free

Auto-tag any photo with AI-generated labels and confidence scores using CLIP ViT, running entirely in your browser. Free, no signup, no upload.

How it works

1Upload any photo by dragging it in or clicking to browse. Supports JPG, PNG, and WebP up to 10 MB.
2On first use the CLIP ViT-Base-Patch32 model (~150 MB) downloads and caches permanently in your browser — subsequent runs are instant.
3Click "Tag Image" — CLIP analyses the photo and scores it against 15 descriptive labels (person, animal, food, vehicle, nature, building, technology, and more).
4The result shows each matched tag as a coloured badge with a thin confidence bar and percentage score, sorted from most to least confident.
5Click "Copy tags" to copy all tag labels as a comma-separated list, ready to paste into any metadata field or caption.