Attractiveness test.

Not reliably. A chatbot looks at the photo the way a person would and often gives a different answer each time, or declines. This test measures 17 proportions from 478 landmarks and gives the same numbers for the same photo, with the lines it measured drawn on your face.

Largely, yes. People agree on some things, such as clear skin, symmetry and proportions close to the common range, but taste, culture and expression matter a lot. That is why this test reports how typical and balanced your proportions are, not how attractive you are.

Yes. The free test runs in your browser, so your photo is not uploaded or saved on our servers, and you don’t need an email or account. Only if you later choose to chat with Mia, the AI stylist, is a photo you pick sent, and you can delete it any time.

No test can tell you that, and a low number does not mean it. The score describes how close the proportions in one photo sit to the common range of adult faces; the camera angle, light and expression alone can move it. What changes how people see you most is in your hands: haircut, brows, glasses, colors and how you are photographed. Start with your strongest part, then use the free tools or Mia, the AI stylist, to build on it.

About 7.5 out of 10 is typical on FaceRater: half of the 2,560 reference faces score below it, and 8.4 or more is the top tenth. The percentile next to your score says how many reference faces you are ahead of, which is easier to read than the number alone. Other sites use other scales, so compare scores only within one tool.

PrettyScale and similar sites score how closely marked points match fixed ideals such as the golden ratio, and give one number. FaceRater finds 478 landmarks automatically, compares 17 proportions with 2,560 real adult faces from a set balanced across seven groups, and shows every measurement with the line it came from. No photo test measures attractiveness itself; pick the one whose method you can check.

A picture test can describe the face visible in that one image. Start with the landmark map: do the points actually follow your eyes, mouth and outline? Then open the five parts beside your harmony score to see which proportions drive it. This is more informative than treating one number as how everyone sees you. A photo cannot capture your movement, personality or the setting in which someone meets you.

Adults can use the same free photo test without selecting a gender. Each proportion is compared with the closest group in a reference set balanced across women and men and seven perceived race groups, so a proportion that is common for any of them is not marked as unusual. There are no separate attractiveness ladders. Use an unobstructed frontal photo and judge styling ideas by your own preferences.

Different tools may measure different features, use different weights or produce a subjective model response. Even a shared 0–10 scale does not make their scores equivalent. To compare two results meaningfully, use the same tool, similar photo conditions and the underlying measurements. Paying for a FaceRater report does not recalculate your free score.

Yes. Your harmony score, percentile, five part scores, 17 proportions, ten more measurements and the measurement map are free, and the result stays saved on this device. You can also download a share card or compare two photos in Mog Battle. The AI stylist starts with 10 free credits; the full style report uses 200 credits, and packs start at $4.99.

We correct head roll, turn and nod using the landmark depth, then measure 17 proportions. Each is compared with 2,560 front-facing adult photos from FairFace, a public research set: values near the common range score well, and smaller left–right differences score well. The proportions form five parts (thirds 25%, fifths 25%, eyes 20%, jaw and chin 20%, symmetry 10%), and the combined result is ranked against the same photos. On the 10-point scale a typical face reads 7.5; the top tenth reaches 8.4. We do not use the golden ratio. The full method is on our methodology page.

Camera distance changes apparent proportions. Head turn, a raised eyebrow, a smile, glasses or uneven shadows can shift landmark positions. We correct small head turns and mark the score as approximate when the angle or expression may move it, and we ask for a new photo when the head is clearly turned, the mouth is open or the eyes are closed. Compare photos at eye level, with a relaxed expression and similar lighting.

The local machine-learning model detects facial landmarks. Our own fixed method turns those positions into a harmony score: how close your proportions are to the common range of real adult faces. No AI model votes on your looks. Being typical is not the same as being attractive, so the score is a starting point for styling, not a verdict.

No. Your free result needs no email or account, and it stays saved on this device. To go further, sign in to your AI stylist with Google or a one-time email link; new accounts get 10 free credits. There is no password.

No. The free scan runs in your browser and the photo stays on your device. The result is saved in this browser, so you can come back to it without scanning again.

Use a clear, unfiltered, front-facing image of one adult, with the full face visible. Keep the camera level with your eyes, your head level and your expression relaxed. Avoid wide-angle close-ups. Use even light and remove anything hiding the eyes or jawline.

A hairstyle can change how a face is framed and how a look feels, but it does not change the underlying eye or nose landmarks. Use styling previews to choose a look you like; do not expect a haircut to guarantee a higher face-rating score.