ChatGPT images carry a hidden label
Images created in ChatGPT include Content Credentials (C2PA) — a signed record inside the file saying it was generated by OpenAI. When you upload an original ChatGPT image here, we read that record and confirm it instantly. Downloaded files are often named “ChatGPT Image …”, which we also recognise.
Screenshots, edits and copies shared through social apps lose that label. For those, our pixel detectors look for the patterns ChatGPT’s image model leaves behind: smooth, evenly lit surfaces, a slightly warm colour cast, very clean typography, and over-tidy compositions.
How to spot a ChatGPT image by eye
- A faint yellow/sepia tint, especially in “photo” styles.
- Text that is unusually crisp and correct for an AI image — but with odd kerning or a misspelled small word.
- Common sizes: 1024×1024, 1024×1536 or 1536×1024 pixels.
- Studio-perfect lighting on everyday scenes and very symmetrical layouts.
Frequently asked questions
Can you still detect a ChatGPT image if it was screenshotted?
Often, yes — but with less certainty. The label is gone, so the verdict comes from the pixel detectors and shows a confidence level instead of “confirmed”.
Does this detect DALL·E 3 images too?
Yes. DALL·E 3 images (in ChatGPT, Bing or the API) are covered by the same checks.
Is IsItMadeByAI really free?
Yes. Every scan is free, there is no sign-up, no credit card and no watermark on your results. The site runs on open-source detection models, which is how we keep it free.
Do you keep my photos, videos or text?
No. Files are analyzed in memory and thrown away as soon as your result is ready. Videos never leave your device — your browser pulls out a few frames and only those are checked. Nothing is published unless you tap Share, and then only a small preview is saved.