ElevenMAGAZINE / ISS. 11.04
FEATURE//THE TOOL

Celebrity look alike finder.

One photo. Five matches. Two seconds. A celebrity look alike finder built on real face-recognition math, scored by cosine distance, and engineered to delete your image the moment the work is done.
Skip to the tool
START HERE

Run a match. Right now.

Drop a photo here, or click to browse
JPG // PNG // HEIC — UP TO 5.0 MB
Photo is analyzed and deleted. Never stored unless you opt in.

The internet has been asking the same question since the first photo booth: who, exactly, do I look like? Most celebrity look alike apps answer with a coin flip dressed up in a progress bar. We wanted something more honest. Eleven’s celebrity look alike finder runs your photo through ArcFace, the same family of models that decides whether your passport matches your face at the airport, and ranks you against fifty hand-picked celebrities by mathematical similarity rather than vibes. What comes back is a card with a top match, four runners-up, and a breakdown of which features the model actually noticed. No account required. No tracking pixels. No upsell on the result page. Just a celebrity face match, reasoned out from the geometry, served in roughly the time it takes to pour a glass of water.

Two faces become two vectors in 512 dimensions. The closer the angle between them, the closer the resemblance. There is no closer reading than that.

How it works.

A photo goes in. The model does the math. A verdict comes out. Three moves, in order.

Upload. Drop a single front-facing image into the panel. JPG, PNG, or HEIC, up to five megabytes. The model wants clear light and visible eyes; anything portrait-grade will do. EXIF metadata, the camera details and geolocation tag your phone quietly attached, is stripped the moment the file lands on our server. No account. No tracking. You confirm two things: that you’re eighteen, and that the face is yours.

Analyze. Once the upload completes, your face is cropped and passed through Buffalo_L, an open-source ArcFace implementation built by InsightFace. It converts your features into a 512-dimensional vector: a list of 512 numbers that, taken together, describe your face the way coordinates describe a point on a map. Those numbers are then compared to the celebrity vectors stored in pgvector, our Postgres-based similarity index. The comparison itself is one equation, repeated fifty times. Compute time runs under 200 milliseconds. The full upload-to-result loop takes around two seconds on a normal connection.

Result. What you get back is a verdict, not a guess. Top match in large type. Four runners-up beneath it. A similarity score for each. Then the breakdown grid: face shape, eyebrow score, smile score, all derived from a 106-point landmark map of your features. Share the card. Opt the photo into the public gallery for that celebrity. Or close the tab and the photo is gone. The default is gone.

The science.

There are two ways a computer can look at a face. It can classify, which means picking from a fixed menu: male or female, smiling or not, this celebrity or that one. Or it can recognize, which means converting the face into a kind of fingerprint and comparing fingerprints against each other. Recognition is the better tool. It is what opens your phone. It is what the e-gate at Heathrow uses to wave you through without a stamp. And it is what we use, turned inward toward fame instead of identity.

The fingerprint, in our case, is a 512-dimensional embedding. Think of it as the face’s coordinates in a very high-dimensional space. Each dimension captures something, perhaps the angle of a brow, the ratio of a jawline, the curve of a lip, without a human ever having to label what it captures. The model learned those features on its own, from millions of training images, by being asked over and over: same person, or different? After enough rounds, it figured out which arrangements of pixels tend to belong to the same face.

To compare two faces, we use cosine similarity. It is, simply, the angle between the two vectors. Identical twins read close to zero. Strangers read close to ninety degrees. Your celebrity look alike, the one the model says you most resemble, is whichever of the fifty in our library lands at the smallest angle to you.

Why fifty celebrities and not five thousand? Because curation matters more than scale at this stage. We picked faces that read distinctly across the model’s geometry, span the major demographic combinations, and hold up under the breakdown grid. A bigger library would dilute the experience without making it more honest. A celebrity face match is only as good as the faces it can match against.

03 // RUN IT

Try it.

Below is the tool itself, the same engine described in the rest of this article, available now, no signup, no email gate. Drop in a photo, tick the two boxes, and the model will return your celebrity doppelganger in roughly two seconds. The result card is yours to keep, share, or delete. Nothing is stored unless you ask us to store it.

EMBEDDED // 002

Match me. Now.

Drop a photo here, or click to browse
JPG // PNG // HEIC — UP TO 5.0 MB
Photo is analyzed and deleted. Never stored unless you opt in.

Privacy.

Most celebrity look alike apps treat your photo like inventory. We treat it like a courier package. It arrives, it gets read, it gets shredded.

When you upload, your image is processed in memory: the embedding is computed, the comparison is run, and the original file is deleted within seconds of your verdict landing on screen. EXIF data is stripped on arrival, so the camera serial number and GPS coordinates your phone tucked into the file never reach our database. There are no accounts to create, no fingerprinting scripts, no third-party trackers measuring how long you hover.

We do not sell photos. We do not train models on them. We do not share them with any third party. The only exception is the public gallery: on the result page, you can opt in to appear on the celebrity’s look-alike page. Opt-in is opt-in. The toggle is off by default. Photos that pass it are run through a Gemini-based moderation pipeline that screens for minors, NSFW content, multi-subject frames, and low-quality images before publication. Anything ambiguous gets rejected.

Visitors from the EU, UK, and EEA can read this page but cannot upload. Biometric data is regulated tightly under GDPR and the UK biometrics regime, and we will not process it without a finalized DPA, DPIA, and lawful basis. The geo-block is a feature, not a workaround. We would rather block traffic than mishandle data.

05 // QUESTIONS

Frequently asked.

How accurate is the matching?

The model ranks similarity, not identity. It does not decide whether you and the matched celebrity are the same person; it decides which celebrity in our library sits closest to your face geometry in 512-dimensional space. The number you see is a similarity score, not a probability. A high score means the model thinks the bone structure, feature ratios, and contours line up. It does not mean you are anyone’s twin in any legal or biological sense.

What happens to my photo?

It gets read, scored, and shredded. Your image is held only long enough to extract the face vector, run the comparison, and render your result. EXIF metadata is stripped on upload. The original file is deleted within seconds. The vector itself is discarded after the match unless you opt the photo into the public gallery, in which case the moderated, stripped image is what gets stored, never the raw upload.

Can I see other people’s matches?

You can see the public gallery for any celebrity in our library: the photos that other visitors have actively opted to share, after they passed moderation. You will not see anyone’s private match. Anonymous browsers can scroll the galleries without uploading anything themselves. Each gallery is ranked by similarity score, so the top of the page is always the closest celebrity look alike currently on file for that subject.

Why am I getting weird matches?

Three usual culprits: poor lighting, extreme angles, or a face partially obscured. ArcFace works best on clean, front-facing portraits with both eyes visible. Sunglasses, heavy filters, side profiles, and group shots throw the geometry off. If your match feels off, retry with a brighter, straighter photo. The model is honest about the input. It tells you what your photo looks like, not what you wish it looked like.

Can I match against a specific celebrity?

Only the fifty currently in our curated library. The model can rank you only against faces it already has vectors for. We chose those fifty deliberately, prioritizing distinct face geometries across demographics so the match feels meaningful rather than statistical. The library is the library. There is no premium tier, no paid expansion, no API. The full list is browsable on our celebrities page.

Is this free?

Yes, completely free!

Will my photo end up in the celebrity’s gallery?

Only if you opt in on the result page, and only if the moderation pipeline approves it. The default state is off. You will see a clearly labeled toggle on your verdict screen. Leave it untouched and the photo is deleted. Switch it on and the image goes through automated review for minors, NSFW content, multiple subjects, and image quality before it ever appears in a public gallery. Opt-in is opt-in.

Why are there only 50 celebrities?

Because the size of the library is doing real work. Fifty curated faces let us guarantee that every match is against someone the model handles cleanly, every gallery has enough volume to feel alive, and every breakdown score is calibrated against a face we already know how the model reads. A larger library is a different product, with different trade-offs, and we are not in a hurry to build the wrong version of it.

Try the tool.

You came here for a celebrity look alike finder, not a manifesto. The panel is upstairs. Drop in one photo, tick the two consent boxes, and you will have your verdict, top match, four runners-up, breakdown grid, in about two seconds. No signup. No payment. No retention you did not ask for. Find your celebrity twin and decide for yourself whether the math knows you better than the mirror does.