About Mavqen

Every score comes from Sigil, our rating engine. It places 77 anatomical landmarks — 49 on your front photo, 28 on your side profile — and then asks you to check and correct each one against a definition diagram and a reference photograph. A misplaced jaw point moves several measurements at once, so that step is where most of the accuracy comes from.

From the corrected points Sigil computes distances, ratios and angles — 44 of them in Harmony alone — and reads the face across five pillars: Harmony, Features, Appeal, Dimorphism and Angularity. Harmony is pure geometry. The other four partly rely on a vision model reading the photograph, and every value produced that way is labelled assessed rather than measured, so you can see which numbers are which.

The pillars combine into one score out of 10, weighted differently for men and women. 5 is the average and the scale is deliberately non-linear. It is never a ranking against other users: there are no leaderboards, and your number does not move because someone else scanned.

Face photos are biometric data, and we treat them that way. Your analysis photos are stored privately against your account. They are never public, never shown to other users and never sold, and there is no public profile, feed or leaderboard to put them on.

Mavqen reads your front and side photos together, lets you correct the landmarks the score is built on, and tells you which parts of it are geometry and which are judgement. The analysis then feeds a protocol of what to change — ranked by how much each change is expected to move your score, with non-surgical routes alongside surgical ones — a 2D simulation of the result, and, on Pro, a 3D head built from your own photos.

Questions, corrections, data requests or partnerships: support@mavqen.io.