Free face test · private by default

Free AI Face Shape Detector

Use a free AI face shape detector to compare face length, width, forehead, cheeks, and jaw, then see matched hair and frame guidance.

Upload a photo

Drag and drop or choose a file

JPG, PNG, WEBP up to 10MB

Your photo never leaves your device for the free scan

How the result works

What this face shape detector actually measures

FaceRating.me separates the visible measurements from the interpretation, so you can see what produced the result instead of receiving a mystery score.

face length-to-width ratio

forehead, cheek and jaw width

visible taper from cheeks to chin

What you get

A shape classification with hair and eyewear directions that balance the outline.

Face outline explained

How a face shape detector classifies a photo

A face shape detector compares the outline relationships visible in one front-facing image. It looks at length, cheek width, forehead width, and jaw width, then assigns the closest of six practical labels: oval, round, square, oblong, heart, or diamond.

Those names are styling shortcuts, not fixed biological types. Real faces occupy a continuous range, so a person can sit near the boundary between two categories. The page shows the length-to-width value and a short balance direction to keep the label in context.

The result can help narrow hairstyle and eyewear options, but it cannot choose a style for you. Hair texture, density, frame fit, prescription, personal taste, and how a style looks in motion still matter.

Rules-based classification

What the AI face shape detector compares

The landmark model finds the points; a transparent set of proportion rules assigns the label. The page does not ask a generative model to guess from the whole portrait.

Round

length and width are close

A shorter outline with a jaw narrower than the square threshold is classified as round.

Square

shorter outline + wider jaw

A near-equal length-to-width relationship with a relatively broad jaw produces the square label.

Oblong

length ÷ width above 1.62

A clearly longer visible outline is classified as oblong.

Heart

forehead wider than jaw

A forehead span more than 1.14 times the jaw span produces the heart rule when earlier rules do not match.

Diamond

cheeks are the widest area

The cheek span must be clearly wider than both the forehead and the jaw.

Oval

balanced fallback

Photos that do not cross the other rule thresholds are labeled oval. Borderline results deserve a controlled retake.

Better outline detection

How to find your face shape from a photo

  1. 01Pull hair away from the forehead, temples, cheeks, jaw, and chin.
  2. 02Use a straight head position with the camera at eye level and the full outline in frame.
  3. 03Avoid a close wide-angle selfie, a strong smile, or shadows that erase one side of the jaw.
  4. 04Read the proportions with the label. If a retake crosses a category boundary, treat your shape as a blend.
Start with my photo

Style, not a rulebook

Use face-shape advice as a shortlist

Angular frames can contrast a round outline, while rounded frames can soften a square outline; volume placement can create similar visual balance in hair. These are starting points, not restrictions.

A good fit and a style you enjoy matter more than obeying a category. The detector cannot evaluate hair texture, prescription needs, frame bridge fit, or the view from the side.

Face Shape FAQ

Know what the result can — and cannot — tell you

How does the face shape detector work?

The detector maps 478 landmarks and compares visible face length with cheek, forehead, and jaw widths. A rules-based classifier then selects the closest label among oval, round, square, oblong, heart, and diamond.

What face shape do I have?

Upload a straight-on photo with the forehead, cheeks, jaw, and chin visible. The result names the closest category and shows the length-to-width relationship used by the classifier. Many faces sit between categories, so treat the label as a styling shorthand.

What measurements determine face shape?

The main signals are visible face length divided by width, jaw width relative to cheek width, forehead width relative to jaw width, and whether the cheek area is the widest part of the outline.

Can my face shape be a mix of two types?

Yes. Face-shape names divide a continuous range of proportions into convenient categories. A face near a decision boundary can reasonably look oval in one photo and oblong, round, or heart-shaped in another.

Why does hairstyle affect face shape detection?

Hair can hide the forehead, temples, cheeks, or jaw and change the visible outline. Pull hair away from those areas for the scan, then use the classification as one input when considering volume, length, or frame shape.

Which glasses suit my face shape?

The page gives a broad balance direction, such as angular frames for a round outline or rounded frames for a square outline. Fit, bridge width, lens size, prescription, and personal style matter more than a category alone.

Is an AI face shape detector always accurate?

No. It estimates a three-dimensional outline from one two-dimensional photo. Head turn, wide-angle distortion, expression, facial hair, and hidden landmarks can change the classification.

Is my face shape photo private?

The free measurement runs in your browser, so the selected photo is not uploaded for the scan.