Meaw Prompt — Image to Prompt AI

Right-click any image and get the AI prompt that would recreate it — subject, camera, lighting, palette. 7 output formats.

Turn any image on the web into a prompt you can paste straight into Midjourney, Stable Diffusion, Flux, or any other image generator. Right-click an image and Meaw Prompt reads it the way a photographer would — subject, composition, shot size, focal length, depth of field, lighting direction and quality, colour palette, materials, style, era, mood — then writes the prompt that would recreate it. The result streams into a card on the page. Copy it and carry on browsing. SEVEN OUTPUT FORMATS · English, ไทย, 中文, 日本語 — full descriptive prompts · Structured JSON — subject, camera, lighting, palette, negative_prompt, tags · Midjourney — one dense line plus suggested parameters · Stable Diffusion / Flux — weighted positive tags plus a negative line Switch format on a result you already have. The image is never re-uploaded. CAPTURE ANYTHING ON SCREEN Some visuals aren't images you can right-click — canvas elements, video frames, maps, CSS artwork. Press Alt+Shift+P, drag a box around any part of the page, and analyse that instead. TWO WAYS TO PAY — YOUR CHOICE · Credits: sign in and we run the model for you. 20 credits free to start. · Your own API key: Anthropic, any OpenAI-compatible endpoint, or Gemini. Completely free, no account, and nothing is sent to our servers at all. PRIVATE BY DEFAULT An image is sent only when you right-click it or drag a capture box. Meaw Prompt never reads, scans, or crawls pages on its own. Your prompt history and thumbnails stay in your browser and are never uploaded. No ads, no trackers, no analytics. We do not store your images or your prompts, and nothing you submit is used to train a model. SEARCHABLE LOCAL HISTORY Every prompt you generate is kept locally with its thumbnail. Search it from the toolbar icon, copy anything again, delete what you don't want.