Detect your face shape
Free, private, and accurate. The AI maps 478 facial landmark points and classifies your exact face shape with a confidence score. All processing happens on your device. Nothing is stored or uploaded.
Two more ways to analyze
Use photo upload for a quick one-time scan, or enter your measurements manually for a full anthropometric classification with computed ratios.
For best results: good lighting, no glasses, hair pulled back.
We are building personalized product guides by face shape. Enter your email and paste your face shape details below. We will send you a tailored guide the moment it launches.
What determines your face shape
Face shape classification applies geometric proportion analysis to the human face. Rather than relying on visual impression, it uses four to nine anthropometric measurements , precise distances between anatomical landmarks , and calculates the mathematical ratios between them. These ratios are compared against documented thresholds from published craniofacial morphology research to produce a classification.
The two measurements that form the foundation of any classification are face length (trichion to menton , hairline center to chin tip) and cheekbone width (bizygomatic width , outer cheekbone to outer cheekbone). The ratio between these two values, known as R1, is the single most powerful predictor of face shape. An R1 between 1.4 and 1.6 indicates oval. Below 1.15 suggests round or square. Above 1.6 points toward oblong or rectangle.
As additional measurements are provided , forehead width, jaw width, chin width, temple width, cheek height, nose bridge width, and philtrum length , secondary and tertiary ratios refine the classification. The forehead-to-jaw ratio (R2) separates heart from triangle. The cheekbone prominence ratio (R3) identifies diamond. Each additional measurement reduces ambiguity and raises the confidence score, which displays alongside your result.
Not sure of your shape yet? Use the free face shape detector to find out in seconds, then return here for a deeper understanding of your result.
The three primary ratios explained
R1 (face length divided by cheekbone width) determines whether your face reads as long, balanced, or wide. R2 (forehead width divided by jaw width) reveals the proportional relationship between your upper and lower face. A high R2 means a wider forehead relative to jaw , the hallmark of heart shape. A low R2 means the jaw is wider than the forehead , triangle or pear shape. R3 (cheekbone width divided by the average of forehead and jaw width) identifies whether the cheekbones are the dominant width , the defining feature of diamond shape.
When only the two required measurements are available, the algorithm uses R1 alone with a base confidence of 60 percent. Each additional optional measurement contributes a secondary ratio and adds 5 percent to the potential confidence score. A result where multiple ratios independently support the same classification will approach the maximum displayable confidence of 96 percent.
The eight face shapes explained
Academic facial morphology literature recognizes between six and nine primary face shape classifications depending on the study. Our calculator implements the eight most analytically distinct and style-relevant categories, each defined by specific anthropometric ratio thresholds.
How to measure your face shape at home
Manual measurement produces highly accurate results when taken correctly. You need a flexible measuring tape , the kind used for sewing works well. Pull your hair completely away from your face, ensure good lighting, and relax your facial muscles into a neutral expression. Tensing the jaw or raising the brows will alter your measurements.
Take each measurement twice and average the readings. A difference of more than 3mm indicates one reading was likely off-angle. Millimeters are recommended over centimeters or inches for this scale of measurement because they eliminate rounding ambiguity.
Ready with your measurements? Use the manual calculator to get your face shape classification with full computed ratios and a personalized style guide.
Why face shape matters for personal style
Face shape is the primary structural variable that professional hairstylists, eyewear consultants, and makeup artists use to make styling decisions. These fields operate on a principle of visual contrast: styles that differ geometrically from the face create a sense of balance, while styles that reinforce the face's dominant geometry amplify it.
Hairstyle selection
A professional hairstylist assesses face shape before recommending a cut because the same cut produces dramatically different results depending on the face it frames. A layered cut that adds height at the crown and keeps weight below the cheekbones works well for round faces because it creates the illusion of length. The same cut on an oblong face amplifies perceived length to an unflattering degree. Understanding your shape lets you evaluate any recommendation with a structural reason rather than guesswork.
Eyewear selection
Eyewear frames should create visual contrast with your face shape. Round faces benefit from angular or rectangular frames that introduce straight horizontal lines. Square faces benefit from round or oval frames that soften the angular jaw. Oval faces suit virtually any frame geometry because their proportions are already balanced. Knowing your shape turns frame shopping from trial and error into a targeted process.
Makeup contouring
Contour and highlight placement is entirely governed by face shape. The goal is to use shadow to visually recede features and highlight to advance them. Round faces benefit from contouring along the temples and jawline to create the illusion of angles. Square faces benefit from softened, curved contour that reduces jaw sharpness. Heart faces use highlight at the chin to add visual width and balance the wider forehead. Every placement decision references the face's proportional geometry.
Common questions
What is a face shape detector and how does it work?
A face shape detector analyzes the geometric proportions of your face to classify it into one of several defined categories. Our AI-powered detector uses Google MediaPipe Face Landmarker, which maps 478 three-dimensional landmark points across your face in real time. The algorithm extracts key distances between anatomical reference points and computes the ratios between them.
These ratios are compared against anthropometric thresholds from published facial morphology research to produce a classification into one of eight face shapes: oval, round, square, heart, diamond, oblong, rectangle, or triangle. The result includes a confidence score reflecting how clearly the computed ratios indicate one shape over the others.
How accurate is the AI face shape detection?
Accuracy depends on input quality, the number of measurements provided, and how clearly your proportions align with a single classification. Our detector displays a confidence score with every result. This starts at 60 percent with the two required measurements and increases by 5 percent for each additional optional measurement, up to a maximum of 96 percent.
For camera and photo analysis, MediaPipe's 478-point landmark detection produces highly precise measurements under good conditions: even lighting, a frontal face angle, and a clear hairline. Many faces sit near the boundary between two shapes, which is why we display all eight probability scores rather than only the top result.
Is my photo or camera data stored anywhere?
No. All facial analysis operates entirely within your browser using client-side JavaScript and WebAssembly. Your camera stream is captured locally, processed locally by the MediaPipe model running in your browser's memory, and never transmitted to any server. When you upload a photo, the image file is read locally and never sent anywhere.
The MediaPipe AI model is a file your browser fetches once from a public CDN and caches for future visits. No facial images, landmark coordinates, measurement values, or results are stored on our servers at any point. All processing data exists only in your browser's RAM and is cleared the moment you close the page or camera overlay.
What is the most common face shape?
Oval is the most commonly classified face shape, followed by round and square. Studies in craniofacial anthropometry suggest oval accounts for approximately 25 to 30 percent of face shape classifications, though proportions vary by study population and methodology. Round and square each represent roughly 15 to 20 percent, with heart and oblong each around 10 to 15 percent.
Diamond is consistently the rarest classification, typically fewer than 5 percent of measured populations, because it requires cheekbones measurably wider than both the forehead and jaw simultaneously. Rectangle and triangle are also relatively uncommon and are sometimes underrepresented in face shape literature because they were historically merged with square and oval categories.
Can my face shape change over time?
Your underlying bone structure , which determines the positions of your hairline, cheekbones, jaw angles, and chin tip , is established in early adulthood and remains stable for most of your life. The fundamental classification derived from skeletal measurements is unlikely to change significantly.
However, the perceived face shape can shift with changes in soft tissue distribution. Significant weight change affects the jaw and cheek areas most visibly. Fat deposits in the cheeks can make an oval or square face register as more round. Aging-related changes in skin elasticity also alter the visual outline over time. If you have experienced significant weight change or are reassessing after several years, we recommend taking new measurements rather than relying on a previous result.