Image → Prompt
Upload any image and get a detailed recreation prompt formatted for five different image generation tools, plus a negative prompt and a tag list.
When to use it
- Recreate a visual style or scene you've seen and want to reproduce
- Understand why an image generation prompt produces a specific look
- Get a starting prompt for similar images without writing from scratch
- Extract structured visual attributes (lighting, composition, medium) from a reference image
How to use it
Upload an image
Click to upload or drag and drop. Supported: JPEG, PNG, GIF, WebP. The image is analysed by a vision model (Gemini or Groq Llama 4 Scout depending on which key you have configured).
Review the analysis
The tool breaks the image down into structured attributes — subject, composition, lighting, colors, style, camera, medium, mood, and more. Review these before copying the prompts.
Copy the prompt for your target tool
Switch between tabs for FLUX, Stable Diffusion, Midjourney, DALL-E, and Gemini. Each is formatted for that specific tool's prompt conventions. The negative prompt and tag list are in separate sections.
Output formats
FLUXDetailed scene description optimised for FLUX's natural-language style.
Stable DiffusionComma-separated prompt with style keywords and quality tags typical for SD.
MidjourneyNatural-language prompt with --style and --ar parameters.
DALL-EDescriptive prompt following OpenAI's content and style guidance.
GeminiStructured prompt formatted for Gemini Imagen.
A negative prompt and a flat tag list are also generated — useful for style-mixing and LoRA workflows.
Analysis attributes
Via MCP: Use the image_to_prompt tool and pass a local image path. The server reads the file directly — no upload needed. MCP setup →