Search a face.
Find where it lives online.
Upload one photo and get back the public pages and social profiles that face appears on — each with a 50–100 similarity score. It matches the face, not the file, so it finds the same person in a different photo, pose or year.
The search runs on traceaifacescan.app — this page takes you there. Nothing is uploaded from here.
Start a face searchFree preview: you see whether the face was found anywhere before paying for anything.
What people use it for
Catfish checks
Someone's photos look too polished and video calls are always impossible. If the same face turns up under a different name, with more history and more ordinary photos, you have found the real owner.
Your own exposure
Find where your face has been published — old accounts, scraped galleries, sites you never signed up for — so you can ask for removals.
Marketplace and rental trust
A seller or landlord whose profile picture belongs to a stock library or somebody else's account is worth a second look before money moves.
Trust & safety, via API
Flag a new signup whose photo already belongs to another account. Three HTTP calls, examples on GitHub.
How it works
- Upload one clear photo.
One visible face, at least 200×200 px. JPEG, PNG or WebP.
- The face is matched against public web and social content.
Not a hash of the file — the face itself, so a different crop or a different year still matches.
- You get scored links back.
Each result carries a platform, a URL, a handle where there is one, and a 50–100 score.
- Open the ones that matter and judge for yourself.
A score is a signal, not a verdict. Act on strong matches, ignore lone weak ones.
| Score | Tier | Read it as |
|---|---|---|
| 50–69 | weak | Similar face. Could be anyone. |
| 70–79 | possible | Worth opening the link. |
| 80–89 | strong | Very likely the same person. |
| 90–100 | near certain | Same person, or the same source photo. |
Reverse image search is not the same thing
Google Images and TinEye look for that file — the same picture, republished. They are excellent at it and they are free. But a catfish rarely reuses one file; they take somebody's whole gallery. A face search looks for that person, across photos the file-matcher has never seen together. Run both: the image search first, because it costs nothing, then the face search for what it misses.
Questions
Is this free?
Running a scan without credits returns a free preview: you learn whether the face produced hits, not where. Seeing the links costs one credit, bought in packs on the site. No subscription, and credits do not expire.
What happens to the photo I upload?
Nothing is uploaded from this page — it is a landing page. On traceaifacescan.app the photo is deleted on the schedule shown with each scan, and you can delete a scan and its images immediately. No face is kept as a permanent biometric template for later lookup. Read the privacy policy before uploading anybody's photo.
Will it tell me who someone is?
No, and it is not built to. You get public URLs and similarity scores — never names, addresses or personal records. Twins, siblings, look-alikes, heavy filters and AI-generated portraits all score high. A match means this face has been published here before, which is enough to catch a stolen profile picture and not enough to accuse anybody of anything.
Can I look up a stranger?
Don't. Identifying, locating, monitoring or harassing people is forbidden by the terms, and biometric processing is regulated in many places. The legitimate question is narrow: is this profile who it claims to be, and where do my own photos appear?
How is this different from PimEyes or FaceCheck.ID?
PimEyes has the stronger consumer product — alerts and a removal service — on a subscription, with no self-serve API. FaceCheck.ID is cheaper per search in bulk, but its credits are bought with cryptocurrency and they expire. Trace is pay-as-you-go by card with an OpenAPI-documented API and a free preview before you pay. The full comparison is on the site, written by us, so read it accordingly.
Is there an API?
Yes. One POST with a photo, poll for the result, read back scored profile links. OpenAPI spec and Python, Node and curl examples on GitHub, also listed on RapidAPI and as an Apify Actor for running a whole list of photos without writing code.