Start with one front photo.
The free scan needs just this one. Add more views below and they go into your Complete report — the more we can see, the more thorough the report.
What you get back: every part of your face measured against people like you and scored, how close you are to your own potential, and what would close the gap. For the best result: even light, plain wall, camera at eye height.
Facial norms differ, so this changes every percentile. Nothing to do with how you identify — pick the tables you want to be read against.
Take every photo with your glasses off — the front one and all the views. Frames sit on the eye landmarks and the lenses bend what is behind them. Say yes and we ask for one extra photo with them on, just for the frames read.
A photograph cannot tell curl from texture reliably, and the haircut the plan suggests — and the picture of it — is cut into the hair you actually have. Leave it blank if you are not sure.
Next: the capture check. The Complete report then asks for seven more photographs and guides you through each one, because whole sections of the report are read from them and cannot be read from a front photo: your profiles for ears and chin projection, the hair-back shot for your real facial thirds and hairline shape, a skin close-up for the skin map, both ¾ views for jaw taper and neckline, and your smile. Two minutes, still on your device.
We need a couple more photos
of you.
Your face is not flat. How your jaw runs, how far your ears stand off your head, where your hairline actually sits, what your smile does — none of it is visible straight on, and none of it is guessed. Add the angle and it gets measured; leave it out and that section of your report says so. Take them the same way you took the first one; each card shows you the angle. The list on the right ticks off as you go.
Glasses off in every one of these photos — both ¾ views, both profiles, the smile, the skin close-up and the hair shots. Frames sit on the eye landmarks and the lenses distort what is behind them. The one exception is the last card, which asks for a single photo with them on so the report can call the frame shape for your face.
Capture quality
Which reference should we compare against?
Every value is placed on a distribution for the population and norms you choose. We ask rather than guess it from your face, because guessing is less accurate.
Pick whichever you consider your own, or "pooled" if none fits. About the norm tables →
Facial norms differ between men and women, so this changes every percentile. It has nothing to do with how you identify — pick the tables you want to be read against.
Filled in from last time on this device — change anything that's different.
Stays on your device. Height and weight let the report say plainly whether facial adiposity cues line up with BMI; age helps read the skin metrics.
Your report opens on this page, not in your inbox. The address is so we can reach you about it. Two emails at most, and every one has a one-click unsubscribe.
Please enter a valid email to see your report.
Facial measurement report
Methodology . Every value carries a 1σ uncertainty propagated from landmark localisation error. Reference: . Fingerprint .
Your
report
- Compared against
- · change
- Capture
- Photo
- Fingerprint
- …
- Method
- Canthion · MediaPipe Face Landmarker · every value ± 1σ
Your landmark map
478 points · your silhouette · the widths and the midline every proportion is built fromSymmetry
your left and right halves — mirrored · a reference, not a goalYour left side doubled, your right side doubled, and the true photograph between them. Neither composite is a recommendation and neither is "the better half" — they exist so you can see which side carries which trait. Your measured asymmetry is in the numbers above; almost nobody is symmetric, and the amount that reads to another person is far smaller than these composites suggest.
Your recommended looks — AI mock-ups
Complete tier · the plan's grooming, rendered on your own photo · drag each to compareGrooming & styling plan
tailored to your measurements · non-surgical · evidence-gradedKeeping this report
save · share · come backHow do I download or print it?
How do I come back to it later?
How do I share it?
I paid — where's my report?
How to read these numbers
what they describe, and how they were takenProtocol
evidence-graded · three tiersWhat your angles are read for
method stated for every oneYour ¾ and profile views carry real measurements: jaw definition and taper, the dorsum line, chin–lip direction, the neckline, the ears — each read with its method stated and its error band printed, so you can see exactly how it was arrived at. What they are never turned into is a clinical number a photograph cannot support: a gonial angle, an E-line, a chin projection against a bony plane, a millimetre. Those need an X-ray or a calibrated rig, and inventing them is what makes a report unfalsifiable. Every read, and how it is measured →
Not medical advice. Descriptive measurements against published population averages; not a diagnosis; not a substitute for a licensed clinician. Where procedures are named, discuss with a licensed provider.
Not ready to upload? Every measurement this report takes is explained first: the formula, the population norms and the source for each, free.