What model is used for captioning?

Florence-2 Base by Microsoft, running via @huggingface/transformers (transformers.js). Florence-2 is a multimodal model that understands images and generates natural language descriptions.

Why is the first load slow?

Florence-2 Base is a ~250 MB model. It downloads once, then is permanently cached in your browser — so every subsequent caption is generated in seconds without any additional download.

Is my image uploaded anywhere?

No. The model runs entirely in your browser using WebAssembly. Your images stay on your device.

What is the difference between caption types?

Short Caption gives a brief one-sentence description. Detailed Caption provides a longer, richer description with more context. Alt Text is optimised for web accessibility — concise and descriptive, following WCAG guidelines.

Can I use captions for SEO?

Yes — good alt text improves image SEO and accessibility. Google uses alt text to understand what an image shows. This tool can quickly generate a starting point, though you may want to refine the output for specific SEO keywords.

AI Image Captioner

Auto-generate alt text and descriptions for images using Florence-2, running entirely in your browser. Perfect for accessibility, social media, and content workflows. Free, no upload.

How it works

1Upload any photo by dragging it in or clicking to browse. Supports JPG, PNG, and WebP.
2On first use the Florence-2 Base model (~250 MB) downloads and caches in your browser. This is a one-time download.
3Choose a caption style: Short Caption, Detailed Caption, or Alt Text (optimised for web accessibility).
4Click "Generate Caption" — Florence-2 runs locally in your browser and generates a natural-language description.
5Copy the caption to your clipboard with one click. Use it as alt text, social post, or content description.