You might also like

FAQ

The image is read and downscaled inside your browser. Only the downscaled image data is sent directly to the AI provider you selected — BoxTool never receives or stores it.

You need a vision-capable model: Google Gemini, Groq (Llama 4), OpenRouter with a vision model, or a custom endpoint that supports images.

About this tool

Good alt text helps search engines understand your image and supports screen-reader users. This tool produces several alt text variants (short, detailed, SEO-style) plus captions and hashtags for blogs, product pages and social posts. Review the suggestions and adjust them to the real context of the image.