Reprompt — Image to Prompt, Instantly

About

Right-click any image on the web, hover it and hit the scan button, or capture a specific area of the page — Reprompt analyzes the visual details for you. It describes the subject, composition, style, lighting, colors, materials, mood, and other image details, then turns them into structured prompts you can use for AI image generation.

Reprompt is designed for artists, designers, AI creators, moodboard builders, and anyone who wants to understand or recreate the look of an image more accurately.

KEY FEATURES

• Analyze images directly from web pages — hover scan or right-click
• Capture a custom screenshot area and generate a prompt from it
• Detailed prompts in English, Chinese, and Japanese
• Full view: structured breakdown (subject, composition, lighting, color & materials, style) plus a tuned negative prompt
• JSON view for pipelines and power users
• One-click copy for use in any AI image tool (GPT Image, Nano Banana, Midjourney, Flux)
• Local history of your last 20 analyzed images and prompts, with thumbnails
• Draggable glass card that collapses into a compact pill with an overlay on/off toggle
• Use built-in credits or connect your own custom API (any OpenAI-compatible endpoint: Base URL, key, and model of your choice)
• Works across image inspiration and creative websites