Remove Stable Diffusion metadata from your exports
Automatic1111, Forge and ComfyUI store the complete recipe in PNG text chunks: prompt, negative prompt, sampler, CFG scale, seed, model hash. ComfyUI can embed the whole workflow graph as JSON. This is the largest and most revealing metadata payload of any generator.
What lands in the file
Run the checker on a fresh export to see exactly what your version writes. Generators change what they embed between releases, and a page cannot stay accurate about that forever - your own file can.
Nothing is uploaded
Every byte is parsed in this tab. There is no server to send a file to, so there is nothing to trust us with.
Lossless by construction
JPEG, PNG, WebP and MP4 are rewritten at the container level. The image and video data is copied byte for byte.
You see what was found
Every file is inspected before and after, and each finding is named with its size. No silent operations.
No account, no limit
Fifty files at a time, any size your browser can hold, free. No sign-up wall in front of the tool.
Clean last, always
Every export step is a chance for metadata to come back. Upscalers, background removers, schedulers and design tools all rewrite the file on their way out.
Stable Diffusion questions
What metadata does Stable Diffusion write?
PNG tEXt / iTXt chunks, Full generation parameters. Automatic1111, Forge and ComfyUI store the complete recipe in PNG text chunks: prompt, negative prompt, sampler, CFG scale, seed, model hash. ComfyUI can embed the whole workflow graph as JSON. This is the largest and most revealing metadata payload of any generator.
Does cleaning a Stable Diffusion export change the image?
No. JPEG, PNG, WebP and MP4 are rewritten at the container level and the compressed data is copied byte for byte, so the decoded result is identical.
Will removing the metadata stop the AI label?
It removes the metadata-driven trigger, which is what raises the label in the great majority of cases. It has no effect on pixel watermarks or visual classifiers, and we do not claim otherwise.
Where should I run this in a Stable Diffusion workflow?
At the very end, on the file you are about to upload. Any editor or export step after cleaning can write fresh metadata straight back in.
What each generator leaves in your file
Different tools write different things. Knowing which one produced the export tells you what to look for.
ChatGPT Images
DALL-E
Midjourney
Adobe Firefly
Photoshop Generative Fill
Google Gemini / Imagen
Canva AI
Flux
Leonardo AI
Ideogram
Grok
Sora
Runway
Check a file before you post it
The inspector is read-only and takes a second. If it comes back clear and a label still appears, the cause is not in the metadata -- and you will know that instead of guessing.