Image → JSON Analyser

Upload any image to extract a structured JSON description: colours, composition, AI caption, and ready-to-use prompts for Midjourney, DALL·E, and Stable Diffusion. Runs entirely in your browser.

Preview

Drop an image here

JPG, PNG, WEBP, GIF. Any size. Select multiple for bulk.

Free & private. AI captioning uses a local WebAssembly model (~250MB, cached after first download). Nothing is uploaded to any server; everything runs in your browser.
Loading AI model (first run ~30s)…

Upload an image and click Analyse to see the JSON breakdown, colour palette, and prompt suggestions.