AI Image Forensics
Reverse-engineer any image into a complete recreation blueprint — full prompt, negative prompt, camera, lighting, composition, color palette, and the AI model that fits it best.
Drop an image to reverse-engineer
PNG, JPG, WebP — up to 10MB
What AI image forensics reveals
Some images are too good to ignore — you want to know exactly how they were made. AI image forensics breaks any image down into its full recreation recipe: the positive prompt, the negative prompt, the artistic style, the camera and lens, the lighting setup, the composition, the color palette, and which AI model most likely produced it. Instead of guessing, you get the blueprint.
When to use image forensics
- Recreate a stunning AI image you found, exactly
- Audit your own generations to learn what worked
- Study the prompt structure behind a viral image
- Identify the model and settings before you try to match a look
Frequently asked questions
What is AI image forensics?▾
AI image forensics analyzes an image and reconstructs everything needed to recreate it: the full prompt, negative prompt, style, camera, lighting, composition, color palette, and the most likely AI model used. It is the deepest reverse-engineering tool in the suite.
How is it different from image to prompt?▾
Image to prompt gives you a clean, ready-to-paste prompt. Forensics goes further — it adds the negative prompt, the technical camera/lighting breakdown, the color palette, and a model recommendation, for when you need the complete recipe.
Which models can it detect?▾
It recognises the visual signatures of Midjourney, Flux, Stable Diffusion (SDXL), DALL-E, and Ideogram, and tailors the recreation prompt to the model it identifies.
Is it free?▾
Yes — image forensics is free to try with a daily limit and no signup. Pro raises the limit and adds API access.