Beauty Score Test

The concept of a “beauty score” sounds reductive — can you really reduce a face to a number? The honest answer is: partially. Decades of research in facial perception have identified measurable geometric properties that predict attractiveness ratings across cultures with surprising consistency. Symmetry, proportional harmony, and sexual dimorphism show up in study after study as significant predictors.

Our beauty score quantifies these properties using AI. The system maps 468 facial landmarks in 3D space, then evaluates six structural dimensions. Symmetry measures bilateral mirroring along both the X and Y axes, with corrections for head tilt. Proportions compare over a dozen facial ratios against the golden ratio (1.618:1). Skin clarity evaluates color uniformity, texture smoothness, and blemish density using CIELAB color science.

The scoring model is designed to be fair across demographics. Skin luminosity uses a Gaussian bell curve rather than a linear scale, so both lighter and darker skin tones score equivalently under normal lighting. Nose width proportions include dead zones to reduce ethnic bias. Jawline scoring uses a softened power curve in the range where most real faces land.

Your final beauty score on the 1-10 scale is a weighted composite calibrated against a large distribution. It includes a consistency penalty — large gaps between your strongest and weakest dimensions pull the score down, because research shows that facial harmony (balanced features) matters as much as any single standout trait.

What this tool cannot capture: expression, energy, style, grooming, body language — all of which significantly influence real-world attractiveness judgments. Your beauty score is the structural baseline. Think of it as one useful data point, not the complete picture. The analysis runs entirely on your device. Your photo is never stored or transmitted.