Free AI Detector
Free AI detectors that show their work. Every result comes with the evidence behind it, and when the signals don’t settle the question you get an honest “uncertain” instead of a guess. Check a photo or a video clip below.
Drop images here to check themChoose photos to check
Up to 20 at a time · JPG, PNG, WebP, GIF or TIFF · up to 10 MB each
FreeNo sign-upYour files are never stored
Choose a detector
Each detector has its own page, its own kinds of evidence and its own notes on what it can’t do.
AI Image Detector
Checks photos, screenshots and art for signs of AI generation, using file metadata, pixel forensics and a vision model.
Open the AI Image DetectorAI Video DetectorBeta
A first look at a video clip: frames are taken in your browser and scored one by one by the same vision model as the image detector. In beta.
Open the AI Video Detector
An AI detector should show its work
A bare score asks you to take the detector’s word for it. These detectors follow four habits that let you check the answer yourself.
- Evidence first
- Each result lists what was found: the labels a file carries, how the content behaves and what a model made of it. You can weigh the reasons instead of trusting a single number.
- Uncertain is an answer
- When the signals are weak or point in different directions, the verdict is “uncertain”. A guess dressed up as a percentage is worse than no answer at all.
- Probability, not proof
- A verdict estimates how likely something is to be AI-made. It is a reason to look closer, never a judgment about a person.
- Private by default
- Checks run in your browser when they can. Anything that has to reach our server is processed in memory and discarded, never stored or used for training.
How to read a verdict
Results fall into one of three bands, whichever detector you use.
The evidence points to a camera-made or human-made original, and nothing points the other way.
The evidence is thin or mixed. Look at the details, where the content came from and whether other copies exist before you decide.
There are strong signs of generation, such as an AI label carried by the file or a confident model score.
Next to the verdict, confidence (high, medium or low) tells you how solid the evidence behind it is.
Private by design
Your files stay yours. The detectors are built to check them without keeping them.
- What we keep
- Your files are never stored. To make repeat checks instant, the server may hold a fingerprint of an image or a video frame, and its result, in memory for up to 24 hours. A fingerprint can’t be turned back into the original.
- Where checks run
- Images and videos are read in your browser where they can be: a video file is never uploaded, and only a few sampled frames reach our server when a model is needed. Whatever reaches the server is analyzed in memory and discarded right after.
- What we count
- Requests per IP address, so the service stays fair for everyone. The counters live in memory, expire within 24 hours and aren’t written to logs.
- What we don’t do
- No accounts, no advertising or cross-site tracking cookies, and no selling or sharing of personal data.
What no AI detector can tell you
Treat a result as evidence, not as a verdict on a person, and weigh it against where the content came from.
A score is not proof.
Every detector estimates likelihood. A high score is a reason to look closer, and a low one is not a certificate of authenticity.
Detectors trail new generators.
New AI models keep arriving, and detection is retrained after the fact. Expect misses on the newest tools.
Editing and compression blur the signals.
Filters, resizing, screenshots and repeated re-saving can push a result in either direction.
Never accuse someone on a score alone.
Get independent verification before any decision about a person, an academic, employment or disciplinary one especially.
Guides
Short, sourced explainers on checking content for signs of AI.
How to Tell If an Image Is AI-Generated
Five checks that still work, from the labels inside the file to a reverse image search, and why hunting for glitches no longer does.
Read the guideHow Do AI Detectors Work?
The four ways a detector can tell, from labels and watermarks to classifiers and statistical tests, and why each one is an estimate.
Read the guideAre AI Detectors Accurate?
Independent studies, a regulator’s order and a test on a public benchmark: what detectors get right, where they fail, and how to read an accuracy claim.
Read the guideTurnitin AI Detector: How It Works
What Turnitin’s AI writing score measures, who sees it, what it claims about false positives, and what to do if your work is flagged.
Read the guide
Frequently asked questions
Accuracy, privacy and what to do with a result.
Is there a free AI detector that is accurate?
This one is free, and it shows the evidence behind each verdict so you can judge a result case by case. No detector is perfectly accurate: results are probabilities, and the page for each detector explains where it does well and where it struggles.
Which detectors are available?
The AI image detector and the AI video detector (in beta) are available now. Each gets its own page with its own evidence and its own notes on accuracy, so read the page before relying on a result.
Do I need an account?
No. There is no sign-up, subscription or credit system. Fair-use limits keep the detectors fast for everyone.
Do you store what I check?
No. Images and videos are read in your browser where they can be, and images and video frames that reach our server are analyzed in memory. What you sent is discarded right after. It is never saved or used to train models.
What does “uncertain” mean?
The signals are weak or point in different directions. It is an honest answer, not a failure: look at the evidence, where the content came from and whether other copies exist.
Can an AI detector be wrong?
Yes. Heavy editing, compression, unusual styles and brand-new generators can all mislead a detector. Treat a result as one piece of evidence, and never as the only basis for accusing anyone.
Still have a question? Contact us
Try the free AI detector
Drop in an image or video clip and get a verdict, with its evidence, in seconds. Free, and no sign-up.
Check now