Skip to content

Product

How FaceSearch works

FaceSearch takes one photo, finds the face in it, turns that face into a compact mathematical description, and compares it with descriptions of faces found on publicly accessible web pages. The closest matches come back ranked, each with the page it was found on. A deep search repeats the comparison over a wider set of public sources.

First search free. No account needed. Updated November 5, 2026 by The FaceSearch Team.

The pipeline, step by step

  1. You add a photo

    A JPEG, PNG or WEBP with one face in it, dropped on the search page or in the app. Large images are downscaled in your browser before upload, so a 12-megapixel phone photo does not need to travel in full.

  2. Metadata is stripped and the image re-encoded

    Before anything looks at the face, the file's embedded metadata (camera, time, GPS location) is removed and the image is re-encoded. The search never sees where or when the photo was taken.

  3. The face is found

    The engine locates the face in the frame. One clear, front-facing face gives it the most to work with; the photo tips page explains what helps and what hurts.

  4. The face becomes a description

    The face is converted into a compact set of numbers that summarise its geometry and texture. This description is what gets compared. It is not an image and it cannot be turned back into your photo.

  5. Descriptions are compared and ranked

    Your face's description is measured against the descriptions of faces found on publicly accessible web pages. The closest come first. A normal search covers the core set of sources; a deep search covers more and takes longer.

  6. You see pages, with their sources

    Each result shows the matched image and links to the public page it came from. On the Free plan the top results are visible and the links are locked; paid plans unlock every link. Signed in, the scan is saved to your history.

What a face description is

People often picture a face search as the engine flipping through photos and eyeballing each one. It does not work that way. Every face, yours and the ones found on public pages, is reduced to a list of numbers that captures the relationships between features: the spacing of the eyes, the shape of the jaw, the texture of the skin, hundreds of such measurements compressed together. Two faces are "close" when their lists are close. This is why a search is fast, why it can match a face across different photos, lighting and ages, and also why it sometimes matches a stranger who happens to measure alike; see the doppelgänger effect.

Normal search compared with deep search
Normal searchDeep search
SourcesThe core set of public pagesA wider set of public pages
TimeSecondsLonger; it is doing more comparisons
CreditOne normal creditOne deep search credit
Who can run itAnyone, including a first free search without an accountAccount holders with deep credits (all paid plans include some)
Best forA quick check, a first lookA thorough self-search, a check that came back empty, a takedown sweep

What is stored, and for how long

  • Your uploaded photo is used to run the search. Its metadata is stripped first. It is never sold, never shared with other users, and never used to build a face database of the people who search.
  • Your saved scan, when you are signed in: the search and its results, so you can reopen them without searching again. Delete any scan from your history whenever you like.
  • Your account data: email, plan, credits and billing records needed to run the service. Export it from your account, or delete the account and everything attached to it. The delete your account page lists what is kept for legal and fraud-prevention reasons.
  • Nothing about you is added to the index. The index is built from faces on public web pages. Searching does not put your face in it.

Where the faces in the index come from

From publicly accessible web pages: pages anyone can open without logging in. Private profiles, stories, direct messages, private albums and anything behind a login are not fetched and not indexed. That is a hard boundary, not a setting. It also means that if a photo of you exists only inside a private account, FaceSearch will not find it, and neither will any other face search engine that respects the same boundary.

Why results change over time

The web moves. Pages are published, edited and deleted every day, and the index follows. A search run next month can return pages that did not exist today and lose pages that have since been taken down. A result whose link no longer opens usually means the page changed after it was indexed; reading your results explains how to handle that. For anyone keeping an eye on their footprint, a periodic self-search is the practical approach, since FaceSearch does not offer alerts.

Why some searches take longer

A deep search compares your face against far more pages, so it takes longer by design. A normal search can also take a moment longer when the service is busy or when the photo needs more work to find the face in it. The scanner shows progress while it runs, and a saved scan lets you come back to a finished result rather than wait on the page.

Privacy, in one place

The short version is above; the full version is on privacy by design, which covers retention, export, deletion and who can see your results (only you), and in the privacy policy, which is the legal text. If you would rather not appear in results at all, the remove my photos page explains the free opt-out.

Questions

01Do you keep my photo?

The photo is used to run your search, after its metadata is stripped. Signed in, the scan and its results sit in your history until you delete them; deleting your account removes them all. The photo is never sold and never used to build a face database.

02Do you build a database of users' faces?

No. The index is built from faces on publicly accessible web pages. Searching does not add your face to it, and we do not keep a separate database of the people who search.

03Why does a search sometimes take longer?

A deep search compares your face with many more pages, so it is slower by design. A normal search can take longer when the service is busy or the face is hard to find in the photo. Saved scans let you return to the result instead of waiting.

04Why did my results change since last time?

The index follows the web. Pages appear, change and disappear, so a later search can include new pages and lose deleted ones. A link that no longer opens usually means the page changed after it was indexed.

05What file types and sizes work?

JPEG, PNG and WEBP. Large photos are downscaled in your browser before upload, so phone photos are fine as they are. One clear, front-facing face with a few hundred pixels across it gives the best result.

06Can it see private accounts?

No. Only publicly accessible pages, the ones anyone can open without logging in, are indexed. Private profiles, stories, direct messages and private albums are never fetched.

Your face is out there. Find it first.

Search with a photo

Public pages only. Your photo is never sold and never used to build a face database.

We'd like to use analytics cookies to understand how FaceSearch is used and improve it. Your photos and account pages are never recorded. Privacy policy