AI Attractiveness Test: How Attractive Am I?
Upload a selfie for your free AI score, face shape and symmetry estimate. Explore optional personalized advice in the full report.
Upload a clear photo
Click here or drag and drop your image
Choose one front-facing photo in even lighting
How the attractiveness test works
Upload a clear, front-facing photo
Use even lighting, look straight at the camera, and keep hats, heavy filters and sunglasses out of the frame. The model reads proportion from the image, so anything that hides structure lowers the accuracy of the result.
The model scores your features
The AI estimates twelve feature scores — eyes, eyebrows, lips, symmetry, bone structure, jaw, facial harmony, cheeks, chin, forehead, nose and skin — alongside a separate golden-ratio compliance estimate.
Read your score in context
You get a 1–10 attractiveness score plus the per-feature breakdown, and a percentile so the number means something. The average across 17,032 analyses is 5.59/10.
What your attractiveness score actually means
Most attractiveness tests hand you a number with nothing to compare it against, which is how a perfectly ordinary result ends up feeling like bad news. Here is the real distribution across 17,032 analyses run on this site between January – August 2026.
| Score range | People | Share |
|---|---|---|
| Below 4.0 | 77 | 0.5% |
| 4.0 – 4.9 | 4,230 | 24.8% |
| 5.0 – 5.9 | 8,450 | 49.6% |
| 6.0 – 6.9 | 2,891 | 17% |
| 7.0 – 7.9 | 824 | 4.8% |
| 8.0 – 8.9 | 524 | 3.1% |
| 9.0 + | 36 | 0.2% |
The average is 5.59/10 and the median is 5.3. Almost half of everyone who takes the test lands between 5.0 and 5.9. If you scored a 5, you are not below average — you are the average.
Figures use analyses from January 2026 onward (n=17,032), after the Dec 2025 model recalibration. Earlier analyses used a different scale.
The attractiveness scale, from 1 to 10
The casual 1–10 scale people use in conversation is not the same as a measured one. On a measured scale the top is much narrower than it feels. These are the shares of people scoring at or above each mark:
A 6 puts you in the top 25.1%. A 7 puts you in the top 8.1%. Only 3.3% of people reach an 8. The gap between "average" and "conventionally very attractive" is a lot smaller in points than it is in perception — about two points covers it.
How your score is calculated
The AI estimates twelve feature scores alongside an overall result. The per-feature averages below come from the same 17,032 analyses, along with how strongly each feature tracks the overall score.
| Feature | Average | Tracks overall | Improvable |
|---|---|---|---|
| Jawline | 5.59 | 0.92 | Partly |
| Bone structure | 5.6 | 0.96 | No |
| Skin | 5.3 | 0.87 | Yes |
| Facial harmony | 5.58 | 0.99 | Barely |
| Chin | 5.39 | 0.96 | No |
| Symmetry | 5.62 | 0.95 | Barely |
| Cheeks | 5.42 | 0.96 | Partly |
| Eyes | 5.97 | 0.95 | Partly |
| Nose | 5.36 | 0.97 | No |
| Lips | 5.65 | 0.9 | Partly |
| Eyebrows | 5.94 | 0.9 | Yes |
| Forehead | 5.39 | 0.93 | Partly |
Two things are worth noticing. Every feature correlates strongly with the overall score, so there is no single feature that "carries" a face — attractiveness reads as a whole. And the features that move most between high and low scorers are not the ones people obsess over: jawline, bone structure and skin separate an 8 from a 4 more than the nose does.
Symmetry, the golden ratio, and which one matters
Both get quoted constantly in looksmaxxing circles, usually interchangeably. In our data they behave very differently. Across 17,032 analyses, symmetry correlates with the overall score at 0.95, while golden-ratio compliance correlates at only 0.67.
Symmetry
Average 5.62/10. People scoring 8+ overall average 8.54 on symmetry; people scoring under 5 average 4.74. That is a real and consistent separation.
Golden ratio
Average compliance 77.28%, with the highest observed at 96%. Only 4.4% of faces clear 85% compliance — and clearing it does not reliably produce a high score.
The practical reading: symmetry is worth paying attention to, and golden-ratio compliance is largely a curiosity. If a tool leads with golden-ratio percentages, it is selling you the more impressive-sounding number rather than the more predictive one.
How much can an attractiveness score actually improve?
We measured the gap between each person's current score and the model's estimate of their realistic potential across 17,032 analyses. The average gap is 1.31 points, with a median of 1.3. The tenth percentile is 0.7 and the ninetieth is 1.7 — so almost nobody is sitting on four points of hidden potential.
Headroom is largest at the bottom of the scale, not the top: people scoring under 4.0 average 2.02 points of potential, against 0.32 for those already at 9+. But current and potential correlate at 0.94, and that is the deflating part — the ceiling tracks the starting point. Someone at 4.5 has more room to move than someone at 8, and still does not end up at 8. What moves is grooming, skin, hair, body composition and photography: the inputs that respond to effort. Skin is the clearest example. It is the weakest feature for 38% of people and the strongest for only 11.8%, and it is one of the few things on the list marked genuinely improvable.
What an attractiveness test cannot tell you
It is not objective truth. The model can produce culturally and demographically biased estimates. It cannot tell you how a specific person will feel about your face.
It scores a photo, not a person. Lighting, lens, angle and expression move the result. Two photos of the same face taken minutes apart can differ by a meaningful margin, which is why one bad score is worth re-testing rather than absorbing.
It says nothing about your worth. A measured facial-geometry score is a narrow thing. It does not measure charisma, presence, warmth, or the parts of attraction that decide most real outcomes.
Getting an accurate result
- Face the camera straight on. Angled shots distort the symmetry measurement more than anything else on the list.
- Use soft, even light. Overhead light casts shadows that read as hollow cheeks and heavy brows. Window light facing you is ideal.
- Neutral expression, eyes open. A wide smile changes the proportions of the lower third of the face.
- No filters or beauty mode. Phone beauty filters silently smooth skin and enlarge eyes, which inflates the result and makes it useless as a baseline.
- Keep hair off your face. The forehead and jaw are both measured; hair covering either one costs you accuracy.
- Test more than once. Two or three photos taken in different conditions give you a range, which is far more useful than a single number.
Frequently asked questions
How accurate is an AI attractiveness test?
This is an AI estimate of a photograph, not a validated measurement of attractiveness. Scores can change with lighting, pose and model output. Our published data describes results from submitted photos, not a representative population or a benchmark of human judgments.
What is the average attractiveness score?
5.59/10, with a median of 5.3, across 17,032 analyses. Nearly half of all results (49.6%) fall between 5.0 and 5.9. Scoring a 5 or a 6 is the normal outcome, not a poor one.
Is a 7 out of 10 a good attractiveness score?
Yes — it is roughly the top 8.1%. Only 3.3% of analyses in our sample reach an 8 and just 0.2% reach a 9. People consistently underestimate how high a 7 sits, because the 1–10 scale in casual conversation is not the same as a measured distribution.
Is the attractiveness test free?
The overall score is free with an available scan allowance without signing up. Basic face shape and symmetry are included free; other feature scores and written recommendations require full report access. No credit card is needed for the free score.
Does the test work for men and women?
People can use the tool regardless of gender. We have not published independent accuracy or fairness benchmarks across demographic groups; treat the output as a subjective photo estimate.
Can I improve my attractiveness score?
Somewhat, and by less than the internet suggests. The average gap between a current score and the model's estimate of realistic potential is 1.31 points across 17,032 analyses. Skin is the single most improvable input — it is the weakest feature for 38% of people and it responds to actual effort, unlike bone structure.
Do you keep my photo?
Your photo is processed by our AI providers and saved privately with your report. It is not published. You can delete the scan from Progress or delete your account in settings. See our privacy policy for processing and retention details.
Related tools
Rate My Face
The face rating tool with the full feature breakdown.
AI Face Rater
How the rater works, and what each score band means.
Rate My Photo
Rate a selfie, a portrait or an outfit photo.
Celebrity Look-alike
Match your facial geometry against 3,000+ famous faces.
Hairstyle AI
Preview cuts and colours on your own face.
Am I Pretty?
The same question, answered against the distribution.
Am I Ugly?
Why a score below 5 is far more ordinary than it feels.
Looksmax AI
How much of your score is actually movable.
Research
The studies behind every figure on this page.