Rate your jawline with a free AI jawline rater. A 2D front-photo jaw-angle estimate, symmetry, visible width, and lower-face proportion - computed from 478 facial landmarks and drawn on your photo. No login; the photo stays in your browser.
For a reliable result: keep the lens at eye level, your chin neutral, and both sides of your face equally visible.
Upload a photo
Drag and drop or choose a file
JPG, PNG up to 10MB
Your photo never leaves your device
The jawline score is a 0–100 comparison with four transparent, front-photo reference checks. It is not an attractiveness rating, diagnosis, or population percentile. The bands describe distance from this calculator's reference centers:
90–100 - Very Close
Very close to the center of this calculator's four front-photo reference checks. This describes geometric agreement, not attractiveness or health.
80–89 - Close
Close to the combined angle, width, symmetry, and lower-face references, with a modest difference in one or more components.
70–79 - Moderate Match
A moderate match across the four geometry checks. The component scores explain which measurements are closer or farther away.
60–69 - Mixed Match
Some measurements are close to the calculator references and others differ. The scalar alone does not identify a jawline shape.
Below 60 - Outside Reference
One or more visible-photo measurements are farther from the internal reference centers. This is not a diagnosis or a negative appearance judgment.
Four geometric components use fixed heuristic weights. The angle overlay and left, average, and right estimates are shown with the free result; the report keeps the deeper component interpretation.
The 2D corner formed by the chin, outer-jaw mesh anchor, and side-face mesh anchor on each side. The internal 118°–126° scoring band belongs to this calculator's front-photo coordinate system; it is not a clinical gonial-angle range.
Left-versus-right jaw-anchor distances compared around the detected facial midline. Pose and perspective can affect this visible-photo estimate.
Outer-jaw mesh-anchor width divided by cheekbone-anchor width. It uses the same calibrated 0.76–0.82 reference band and peaked score curve as the report's Jaw-to-Cheekbone Width metric; it is not a direct bigonial measurement.
Nose-base-to-chin distance divided by visible forehead-anchor-to-chin height. It shares the report metric's 0.36–0.42 reference band and score curve.
The clinical gonial angle describes the bend between the mandibular ramus and body at the gonion. That anatomical measurement is not directly visible in an ordinary frontal selfie. This tool instead reports a different, explicitly labeled 2D angle made from visible Face Mesh anchors near the jaw outline. For normal ranges, sharp vs obtuse patterns, and what actually changes the angle, see our full gonial angle guide.
Clinical measure
Anatomical
Uses identified bone or validated anatomical landmarks, often from a lateral cephalogram or suitable 3D imaging.
This calculator
118–126°
Internal scoring band for the front-photo angle formed by the chin, outer-jaw anchor, and side-face anchor. It is not a cephalometric reference range.
How to use it
Compare
Use the number to compare similarly framed photos. Do not use it to infer bone anatomy, health, attractiveness, or a jawline type.
Why the distinction matters: research comparing 2D photographs with direct measurements found poor agreement for bigonial and bizygomatic breadth, and noted that gonion is difficult to locate in a frontal 2D image. Pose and lens perspective add further variation. The result is therefore presented as visible-photo geometry, not a substitute for a clinical measurement.
Front-facing, eye-level camera
Hold the camera at eye level and square to your face. Our pose regression found score movement beginning around 2° yaw or 3° pitch; the page shows an accuracy note when the broader production pose check marks a photo fair or poor.
Neutral expression
Smiling pulls the cheek and jaw landmarks. A relaxed, closed-mouth expression gives the most stable jawline measurement.
Hair off the jaw
Hair covering the jaw outline blocks landmark detection on the affected side, often inflating asymmetry. Tie hair back so the cheek-to-chin line is visible.
Even, soft lighting
Hard shadows under the jaw exaggerate definition; flat lighting hides it. Diffused front lighting (window light works well) gives the cleanest jawline reading.
There is no universal good or bad score. A higher number means the four measurements in this particular front-photo model sit closer to its reference centers; it does not mean a face is more attractive, healthy, or normal. Use the component values to understand the number.
No single gonial angle is perfect across people, measurement methods, and viewing angles. One 2016 internet survey preferred 130° in male profile views, but that study's profile angle is not the same measurement as this tool's front-photo mesh angle. This calculator therefore labels its degree value as a front-photo estimate.
A clinical gonial angle uses anatomical mandibular landmarks, commonly on a lateral cephalogram or validated 3D image. This tool cannot see those bone landmarks. It computes the 2D angle from the chin to two visible Face Mesh anchors on each side, then shows the left, right, and average values without claiming clinical equivalence.
It uses MediaPipe's 478-point surface mesh and deterministic 2D geometry in your browser. The same image produces the same calculations, and the overlay shows the anchors behind the angle. It remains a photo-based estimate: pose, lens perspective, hair, and landmark visibility can change the result.
You can improve the reliability of this photo score with neutral posture, an eye-level camera, even lighting, and an unobstructed jaw outline. Questions about lasting anatomical or medical changes are individual; discuss them with a qualified clinician rather than using a photo score as guidance.
There is not good controlled evidence that mewing reshapes an adult jaw. Changes between photos can reflect head posture, camera angle, expression, lighting, or normal development rather than new bone structure. For bite, breathing, or jaw-development concerns, use advice from a qualified dental or medical professional rather than a photo score.
A lower number can come from any combination of the four components; it cannot be translated into a single label such as soft or square. Photo conditions also matter. Our pose regression showed that even small viewpoint changes can move this 2D score, so compare photos taken with the same straight-on setup.
No. The model uses one transparent set of front-photo reference bands and does not infer or apply a gender-specific attractiveness standard. The result is a geometry comparison, not a demographic norm.
Common descriptive labels include square, tapered, rounded, and angular. They are visual categories rather than diagnoses, and a frontal photo cannot reliably determine profile projection or skeletal position. This tool does not force a shape label from the total score; it reports the underlying angle, width, symmetry, and lower-face checks instead.