AI attractiveness test: how attractive am I?

How the attractiveness score is made

Most online attractiveness tests hold your face up to one formula, such as the golden ratio, and call the match your score. FacialRate does not. Your browser finds 478 landmarks on your photo and measures the spacing between them, all on your device. A machine-learning model then reads those measurements and scores the face. It learned from real human ratings, so how much each facial factor counts is decided by what people actually rated, not by a hand-picked rule.

The model was calibrated against those ratings and tested, which is why the result is more than a guess. It also reads your face more than one way: you see an overall score and tier, but also how each part of your face reads, so you learn where your strengths are instead of staring at a single number.

What your attractiveness result shows

A result is only useful if it explains itself. After one photo, the free face report gives you:

• Your score and tier An overall score with the tier it falls in, so you know what the number means before you read anything into it.
• Your face type One of four face types, with a one-line description and the traits that stand out. Types describe features; none is better than another.
• Your top three highlights The three things the model reads most strongly in your photo, each with a short note.
• Your face shape and jawline Your shape and jawline type, and what each one means.
• Thirds, fifths and 11+ features How your face divides into thirds and fifths, and 11+ features read as strengths or room to grow, with one tip free to try.

What a score can and cannot tell you

A score describes one photo, not you. Light, angle, distance from the lens and expression all change what the model sees, so the same face can score differently on two photos taken a minute apart. That is expected. Treat the result as one reading, and try a second photo before you trust either.

A score also reflects the taste of the people who rated the faces the model learned from. Attractiveness is partly shared and partly personal, and it shifts with culture, age and fashion, so no tool can tell you how attractive you are to one particular person. What it can do is show which features a rating leans on and how your photo reads against them.

No score measures charm, humour, kindness or the way someone looks when they talk and move. It is not a verdict on your worth and must never be used to make a decision about anyone. If a number leaves you feeling worse, close the tab: the guide on how accurate AI face ratings are explains why a low reading rarely means what it seems to.

How to take a photo that reads fairly

The photo matters more than any setting. Four habits give a fairer reading:

1. Face the camera straight on, head level, with your chin neither lifted nor tucked.
2. Use soft, even light from the front, such as a window facing you, not a lamp above you or a window behind you.
3. Keep a relaxed, neutral expression and put the phone at eye level, 1.5 to 2 m away if you can (rear camera, timer): a close lens stretches the nose and the middle of the face.
4. Pull hair back from your forehead and jaw, and take off glasses or anything else that hides your face.

What happens to your photo

Your photo is processed only on your device. Your browser finds the landmarks and takes the measurements itself, and neither your photo nor any face image made from it is ever sent to a server.

The score and a few dimension values can go to the comparison pool, anonymously and without the photo; that is how your ranking is worked out, and you can turn it off in Data & privacy settings. The privacy policy sets out what is sent and what stays on your device.

Read the privacy policy

Keep reading

• Are AI face ratings accurate? First ask: compared with what Five checks that tell a trustworthy face rating from a flattering one.
• What makes a face attractive: three things studies agree on, and their limits What studies agree makes a face attractive, and where they stop agreeing.
• Sample face report A whole sample report, section by section, so you know what you get before you start.