Rate my face: free AI face rating, calibrated to be honest

• Score. An overall 0–100 on a fixed scale that does not drift with the crowd.
• Tier. One of eight bands, from Baseline to Apex.
• PSL equivalent. The 1–8 forum scale, shown beside the score so it means something in the vocabulary you already use. See PSL score.
• Visible traits across five areas. Harmony, eye area, jaw and chin, skin, hair. Each trait gets a band, not a fake measurement, and each is tagged workable or structural.
• Potential. The ceiling reachable without surgery, capped per area in code, and the gap between it and today.

The score, the tier, the PSL equivalent, a share card and a pointer to the area worth working on first are free. The trait breakdown, the weak points with why, the plan and the potential photo are in the paid report.

Most rating apps are tuned for the share button. A result that says you are above average gets posted; one that says you are average does not. So the scale drifts upward until nearly everyone scores well, and the number stops carrying information.

BecomeTen is calibrated the other way. An average, healthy, groomed adult lands between 50 and 60. Scores from 78, the High tier, are reserved for genuinely striking structure, and the top two tiers are rare by design. If your first result is lower than an app you tried before, that is the calibration working. The number describes the photo against a fixed rubric, not your worth.

1. Upload one photo: even light from the front, phone at eye level, neutral expression, hair off the face.
2. The AI rates the photo against the rubric. Every trait gets a band and a workable or structural tag; every potential figure is clamped to its cap before you see it.
3. You see your score, tier and PSL equivalent, free. Decide from there whether you want the report.

No card, no account. Pipeline details: how AI face rating works.

The report is a one-time payment for that scan and stays yours (€6.99, or €1.99 in some countries; in the app, the store price). A re-scan later is €4.99 (€1.99 in those countries). To track change over months, BecomeTen PRO is €9.99 a month or €59.99 a year: 20 scans per rolling 30 days, at most 3 a day, each with its potential photo, plus progress tracking. How to track progress covers the checkpoint protocol.

Asking a chatbot to rate your face out of 10 went viral, and it is easy to see why: it is free and it answers in seconds. The problem is that a general chatbot has no fixed rubric. It is trained to be agreeable, so it tends to be generous, and it can give the same photo a different number when you ask again or phrase the prompt differently. It also cannot tell you how much of its guess came from the lighting, the lens or the angle.

A rating built for the job works the other way round. BecomeTen scores against one published rubric, is calibrated so that an average face lands in the middle rather than at 8, grades the photo before it scores it, and says which areas cost the most. If you want the 1–10 view, the attractiveness scale shows how the 0–100 score maps onto it.

Three things usually explain it, and none is that you are worse than you thought. Calibration: if the last app put you at 8 out of 10, it put nearly everyone there, and a 55 here describes the same face on a scale with room at the top. The photo: overhead light, a lens too close or a tucked chin can cost several points across the eye area and lower third, and the capture grade will say so. Weighting: harmony and the eye area count most, so excellent skin and hair on ordinary structure lands nearer the middle than a flattering rater would suggest.

What to do with it is the reason the product exists. The report splits the score into what can move and what cannot. Most people below 50 are pulled down by one or two workable areas, usually skin or hair, which is where the potential caps are largest. Look at which areas cost the most and start with the cheapest lever that has evidence; the looksmaxxing guide lays that order out.

And if the number lands badly, put the phone down for a day. A rating is one photo through one rubric. It is not a measurement of your face, and not of you.

Every AI rater, this one included, rates a photograph, not a face. A camera flattens three dimensions into two, and how it does so depends on where the light and the lens are and how far away the lens sits. So every photo is graded for light, angle and sharpness before it is scored; a poor capture is flagged, so the score is read with that caveat, not made gentler. Each rule below removes a known distortion.

• Light from the front, even and soft. Overhead light hollows the eye area and exaggerates under-eye shadow; a window in front of you does the opposite.
• Phone at eye level, face square to the lens. A chin pushed out or dropped changes how sharp the jaw reads and shifts the whole lower third; a chin tilted up widens the midface.
• Arm's length or further. A close, wide phone lens enlarges whatever is nearest, usually the nose, and narrows the jaw.
• Neutral expression, hair off the forehead, no glasses. Smiling changes the eye area and cheeks; hair hides the hairline the hair score needs.
• No filters, no beauty mode. Smoothing removes exactly the skin signal the rating reads.

The reasons behind each rule are in photo quality: light, angle and lens.

A free scan processes your photo for the rating and then deletes it; only the text analysis is kept. Photos rated after your free one, and scans waiting for payment, are kept for up to 48 hours so you can unlock them, then deleted with the photo. Nothing is used to train a model and nothing is sold. For a paid scan the photo is stored encrypted in the EU so a later re-scan can be compared against it, and you can delete it whenever you choose. The full rules are in the privacy policy.

BecomeTen is for adults only. A face photo is biometric data, and rating a minor's appearance is not something we will do, so the consent step includes an age confirmation.

Free symmetry and golden-ratio calculators place landmarks on your photo, measure distances between them and compare the result to a template. That is a real measurement, and such tools are honest about being one. Their limit is what they measure: a handful of distances on a 2D projection, against a template that perception research has never validated as a law of beauty.

BecomeTen does something different and says so. It does not measure your face. It gives a structured opinion against a fixed rubric, the way an experienced rater would, across visible traits including the ones a landmark tool cannot see: skin evenness, under-eye condition, hair density, how the jaw reads at your composition. Every structural trait comes back as a directional band, because a number read off a selfie would be false precision.

If you want the measurement, use a calculator. If you want the opinion with its working shown, and a plan that cites only what it can back, that is what this is. The rules are on the methodology page; the attractiveness test page covers what the research supports.