Frequently asked questions
Is AI Image Detector free?
Yes. It is completely free, requires no account, and has no usage limits beyond fair-use rate limiting.
How does the detector work?
It combines five independent forensic techniques: Error Level Analysis (ELA), EXIF metadata inspection, two machine-learning models, frequency-domain AI analysis, and C2PA Content Credentials, then merges them into a single verdict.
Can it definitively prove an image is fake?
No tool can. The result is a probabilistic indicator based on forensic signals. Use it as one input alongside context and other evidence.
What image formats are supported?
JPG, PNG, WebP, BMP and TIFF, up to 10 MB per image.
Do you store my uploads?
No. Uploaded images are processed transiently and automatically deleted within about 20 minutes. We never keep your images.
Why was a real photo flagged?
Heavy compression, screenshots, or strong edits can mimic manipulation patterns. Treat borderline results with the confidence score in mind.
What are Content Credentials (C2PA)?
An industry standard where AI tools can embed signed provenance metadata. When present, we read it to see whether an image self-declares AI generation or editing.