
Landmark alignment: 99%
Eye-width balance: 99.9%
A NUMBER IS ONLY THE START
Your free face rating gives you the measurements. A report connects those details and turns them into suggestions you can actually use.
Read an example report Measured synthetic adult sample · AI styling concepts
Landmark alignment: 99%
Eye-width balance: 99.9%
In this example, both eye widths are similar. The outer corners sit slightly higher than the inner corners. Pose and expression still matter when comparing photos.

44 measurements, one styling concept, six hairstyle images and a four-week plan. See the actual format in the full sample.
YOUR CHOICE, AFTER YOUR FREE RESULT
Two optional reports. One-time prices. Your free score and measurements stay available either way.
Your first look at the numbers.
Understand your rating and know where to start.
We aim to have your report ready within 2 hours. We'll email you when it's ready, so you can safely close this page.
Understand your features, see your options and build a daily plan.
We aim to have your report ready within 2 hours. We'll email you when it's ready, so you can safely close this page.
Start with the 72-Hour Report and upgrade to four weeks later for $10.00. Same photo, same report — no second full-price purchase.
HOW IT WORKS
Start free. A full report is always optional.
Choose a front-facing photo. The free analyzer checks it right in your browser.
See your rating, explore all 10 measurements, and download a share card. No sign-in needed.
Enter your email for a private return link. Explore the preview immediately, then choose a paid report only if you want more.
UNDERSTAND THE RESULT
Stylists describe a face by comparing its length with its width and by where it is widest: forehead, cheekbones or jaw. FaceRater measures those relationships on 478 detected landmarks, levels the eye line and divides each width by cheek width. The mesh top sits on the forehead, not at your hairline, so forehead-led shapes such as heart and diamond are the least certain. When a small change in the photo could flip the answer, we show both shapes.
Read the exact scoring methodFace shapes are styling shorthand. Many faces sit between two of them. These are the rules this detector applies to the measured outline; the thresholds are specific to the landmark mesh, which starts on the forehead rather than at the hairline.
Longer than it is wide, with a jaw that tapers gently from the cheekbones.
Measured as: Length 1.19–1.28× cheek width, with a lower face that tapers or stays balanced.
Length close to its width, full at the cheeks, with a softly curved lower outline.
Measured as: Length under 1.19× cheek width.
A broad lower face that turns more abruptly at the jaw, with similar widths from top to bottom.
Measured as: Length 1.19–1.28× cheek width, with a broad lower face.
Clearly longer than it is wide, with fairly straight sides.
Measured as: Length 1.28× cheek width or more.
Wider through the upper face, tapering to a narrower jaw and chin.
Measured as: A tapering lower face with a clearly wider upper outline.
Widest at the cheekbones, narrower at the temples and jaw.
Measured as: A tapering lower face with temples at or under 0.91× cheek width.
Narrower at the temples, wider through the lower cheeks and jaw.
Measured as: A lower outline wider than the upper outline, without a short length.
YOUR QUESTIONS, EXPLAINED
Compare three things: how long your face is relative to its width, whether the lower face tapers or stays broad, and whether the upper or lower outline is wider. The detector measures all three from a frontal photo and draws the lines on it. To check by hand, pull your hair back, trace your outline on a mirror and compare forehead, cheekbone and jaw widths with the length.
Oval, round, square, oblong (long), heart, diamond and triangle, which is sometimes called pear. They are styling shorthand rather than anatomical categories. Many faces sit between two shapes, which is why your result can name a close second.
Your outline sits near a boundary. If a small, realistic change in one measurement, such as a slightly different camera distance, would produce another shape, we name it. Try a second photo in the same conditions. If the answer flips, take ideas from both shapes.
Only roughly. These shapes depend on forehead and temple width. The landmark model does not find the hairline and hair often covers the temples, so they are the least certain calls. Pull your hair back and use even light; we only name heart or diamond when the measured widths point there clearly.
The soft-tissue outline can change over time, and photos change apparent proportions too. Camera distance, head tilt and expression all move the measured outline. Compare photos taken the same way.
No. Face shape describes an outline for styling choices. FaceRater does not score shapes, and your face rating is calculated separately from this result.
Yes. The scan runs in your browser, so the photo stays on your device. No account or email is needed. A paid report is a separate, optional choice.
Use it as a starting point for hairstyles, glasses and beard shapes. The hairstyle finder uses the same scan to rank short, medium and long cuts for your outline and explains each suggestion.
MAKE SENSE OF YOUR RESULT
Practical answers to the questions that come after a score.
What the number measures, what it leaves out, and where to look next.
4 min read02 COMPARE TWO PHOTOSSeparate changes in a photo from changes in your measured features.
4 min read03 GET A USEFUL SCANA clear front-facing photo makes the measurements easier to interpret.
4 min read