AI IMAGE DETECTOR · FREE PLAN
AI Image Detector: Was This Image Made by AI?
Generated faces, product shots and news pictures now look real. Upload the image to get an explained verdict — probably human, uncertain or probably AI — before you trust, share or act on it.
Open in beta · 10 free credits a month · Files deleted after 30 days
AI Photo Detector
Open in betaWhat you need
Upload an image or paste an image link
JPG, PNG, WebP or HEIC, up to 20 MB.
At a glance
C2PA first
Content credentials and metadata are read before the detector runs; a second model reviews borderline cases.
AnyFact method
3 levels
Probably human, uncertain or probably AI — never a bare percentage — with the reasons and the limits of the method.
AnyFact method
10 credits
Free every month with an account, no credit card. Uploaded files are deleted after 30 days.
AnyFact free plan
Why your eyes are no longer enough
Image generators produce pictures with natural skin, plausible light and readable text. The old tells — six fingers, melted lettering — are becoming rarer with each new model.
A reverse image search does not help much either: a generated image is new, so there is often nothing to match it against.
What does help is combining signals: provenance data when the file carries it, a detector trained on generated images, and a careful reading of what looks wrong — with the limits stated clearly.
Does this sound familiar?
- A profile picture looks a little too perfect.
- A shocking image is spreading fast and nobody knows where it comes from.
- A seller’s product photos look rendered rather than photographed.
- A picture attached to a news story has no clear source.
- You want to know whether an image you are about to publish was generated.
Here's what we check
Upload or paste a link
JPG, PNG, WebP or HEIC, up to 20 MB.
Provenance, then detection
Content credentials and metadata are read first; if they do not settle it, the classifier scores the image and visible anomalies are listed.
A verdict in three levels
Probably human, uncertain or probably AI — never a bare percentage — with a short explanation and the limits of the method.
How to read the result
The signals point to a person. It is not a guarantee: a careful generator or heavy editing can leave few traces.
The signals disagree or are too weak. We say so rather than guess, and list what was found.
The signals point to generation. It is an indication, not proof — check the context before acting.
Sample layout. Real reports name the signals that were checked and the limits of the method.
How to check if an image is AI-generated with the AI image detector
Upload the file or paste a link, then read the verdict together with the signals behind it. A little preparation before you upload makes the result far more useful.
- 01Find the best copy you can. Download the image from the original post instead of taking a screenshot: screenshots and re-uploads drop metadata and add compression.
- 02Upload a JPG, PNG, WebP or HEIC file up to 20 MB, or paste a link to the image.
- 03Let the provenance check run first. Content credentials (C2PA) and EXIF/XMP metadata are read before any model looks at the pixels.
- 04Read the verdict level, then the signals and the plain-language explanation underneath it.
- 05Download the PDF report if you need to keep a record or show your reasoning to someone else.
Each photo analysis uses 1 credit. A free account gets 10 credits a month and needs no credit card.
What does an AI photo detector actually look at?
It looks at three layers of evidence, from the most explicit to the most subtle, and asks for a second opinion only when the score is borderline.
| Layer | What is checked | What a hit means |
|---|---|---|
| Provenance | C2PA content credentials, EXIF and XMP metadata | If the data names a known AI generator, the verdict is probably AI and the reason is stated. |
| Classifier | An open-source AI-image classifier scores the pixels | A score that places the image on the human side, the AI side or in the grey zone. |
| Visible anomalies | Hands, lettering, shadows, reflections, textures | Concrete details you can check yourself and point out to someone else. |
| Second opinion | Claude reviews borderline images | It can move the score only a little and never decides alone. |
The anomaly list is the part you can reuse anywhere. Zoom in on hands, signs and reflections in any picture before you trust it; our guide on how to tell if an image is AI-generated walks through each check by hand.
How to read an AI image checker result
Read the level first, then look at what drove it. The same level can rest on very different evidence, and that changes what you should do next.
| You see | What it usually means | What to do next |
|---|---|---|
| Probably AI, with a provenance reason | The file itself says it was made with a known generator. | Treat it as the most explicit kind of signal: it comes from the file, not from a model’s estimate. |
| Probably AI, from the classifier and anomalies | The pixels look generated and specific details support it. | Check the listed anomalies yourself and look for the original source. |
| Uncertain | The evidence points both ways, often because the image is small, compressed or edited. | Find a better copy or the source. Do not act on this result alone. |
| Probably human | No provenance hit, and the classifier sees a photograph. | Still check the context: a real photo can be old, staged or wrongly captioned. |
When does fake image detection struggle?
Detection is hardest when the evidence has been degraded or mixed. Expect more uncertain results in these cases:
- Screenshots and images re-shared many times, which lose their metadata and gain compression.
- Heavy crops, where the classifier sees only a small part of the picture.
- Real photos edited with AI tools, such as a replaced background or a removed object.
- Images from very recent generators that detectors have seen fewer examples of.
- Illustrations, 3D renders and digital paintings, which differ from photographs in ways a photo-oriented check can misread.
For drawings and digital paintings, the AI art detector page covers the specific pitfalls. When a file carries content credentials, the C2PA checker explains what they contain and what they prove.
Before you check
Is the result proof that a content was made by AI?▼
No. It is an indication, not proof. Detectors can be wrong in both directions, especially on compressed, cropped, edited or very short content. Use the result to decide what to check next, not as a final judgement on anyone.
Why no percentage?▼
A single percentage suggests a precision no detector has. You get one of three levels — probably human, uncertain or probably AI — with the signals behind it and the limits of the method.
How does the analysis work?▼
When the content carries provenance data, such as C2PA content credentials or metadata, we read it first. Then an open-source detector scores the content. When the score is borderline, Claude reviews it as a second opinion, and the result is explained in plain words.
What happens to my files?▼
They are processed to run your analysis — including, for borderline cases, a review by Claude, Anthropic’s model — and deleted after 30 days.
What does it cost?▼
There is a free plan: a free account gets 10 credits every month, with no credit card. Pro ($15 a month) gives 1,000 credits a month; Business ($49 a month) gives unlimited Business Checks and AI checks under a fair-use limit.
If an image has no content credentials, does that mean a person took it?▼
No. Most social networks strip metadata on upload, and many generators add none. Missing credentials prove nothing either way; they are one signal among several.
Can I check an image from social media with a link?▼
Yes, you can paste a link to the image. Keep in mind that social networks usually re-compress pictures and remove their metadata, so the provenance check often finds nothing and the classifier works from a degraded copy. If you can get the original file from the person who posted it, use that instead.
Does the AI image detector work on screenshots?▼
It can analyse them, but a screenshot is a new file: it carries none of the original’s metadata and adds its own compression. Expect more uncertain results, and look for the original image whenever you can.
Can it tell which AI generator made an image?▼
Only when provenance data names it. If the content credentials or metadata mention a known AI generator, the result says so. The classifier score estimates whether an image looks generated; it does not name a tool.
What should I do if I think the result is wrong?▼
Use the “the result is wrong” button on the result page. During the beta, this feedback is how we find the cases where the method needs work. In the meantime, rely on the signals and your own checks rather than on the level alone.
Check this image before you trust it
What you need
Upload an image or paste an image link
Free plan — 10 credits every month, no credit card.
Each result is an indication, not proof.
Files deleted after 30 days.
You might also want to check
AI Art Detector: Was This Artwork Made by AI?
AI art detector for illustrations and digital paintings: check content credentials, score the image with an open-source detector and get an explained verdict.
C2PA Checker: Does This Image Carry Content Credentials?
Check an image for C2PA content credentials and metadata: see which tool created or edited it, and get an AI-detection verdict when no credentials are present.
Deepfake Detector: Is This Video, Photo or Voice Real?
Deepfake detection for videos, photos and voice messages: face swaps, generated clips and cloned voices, checked with open-source detectors and explained. Free plan.