How Attractive Am I? Free AI Test

Upload a photo and get your Face Score instantly: a percentile of how real people tend to rate that photo, checked against human ratings. We also measure 17 facial-geometry features — symmetry, canthal tilt, facial thirds, jaw taper and more — as a map of your proportions, not an attractiveness rating.

17 metrics · NIH-cited landmarks · Private by default
17metrics measured
68facial landmarks
r ≈ 0.8vs 1,566 human-rated faces
100 msfirst-impression window

What makes a face attractive?

The scan measures facial geometry — 17 landmark-based proportions, of which a front photo supports about 15 — plus your Face Score, a percentile validated against averaged human ratings.

SignalWhat we measure
Face ScoreThe attractiveness estimate: a percentile from a face-appearance model, validated against averaged human ratings — r ≈ 0.8 on 1,566 benchmark faces, and r ≈ 0.40 on our own users' phone photos.
Facial SymmetrySymmetrical faces are consistently rated more attractive across cultures. Our AI measures proportional balance across your features.
ProportionsFacial thirds, FWHR (facial width-to-height ratio), midface ratio, nose, lip, philtrum and chin proportions — each read against a reference range.
Eye & Jaw GeometryCanthal tilt, eye shape, brow arch and brow-to-eye distance, and jaw taper (the V-shape from cheekbones to jaw).
Evidence

You have 100 milliseconds

Published facial-attractiveness research (Willis & Todorov, 2006, Psychological Science, DOI:10.1111/j.1467-9280.2006.01750.x) found that people form judgments about your trustworthiness, competence, and attractiveness within 100 milliseconds — before they consciously process what they're seeing.

On dating apps, this means your photo is judged in a fraction of a second. On LinkedIn, recruiters decide whether to click your profile almost instantly. In both cases, the judgment is made on one photo — so the light, the angle and the expression in that photo matter.

Your bone structure is what it is. What you control is the photo — light, camera angle and a relaxed expression decide how clearly it shows your face.

How the attractiveness test works

1

Upload Your Photo

Drop any photo with your face visible. Supports JPG, PNG, WebP, and HEIC. Or use your camera directly.

2

AI Analyzes Your Face

Our AI detects 68 facial landmarks and measures 17 facial-geometry features against reference ranges.

3

Get Your Score & Tips

See your Face Score percentile and your strongest metric. The optional report scores every metric your photo supports.

Why RealSmileWhat it means
Private by defaultPhotos are processed in memory and deleted right after analysis. By default nothing is kept, nothing is used for training, and no other person sees your photo — saving one is always your choice.
Instant ResultsGet your attractiveness score in under 5 seconds. No waiting for human votes, no queues, no credits needed.
Actionable FeedbackMost tests just give a number. We tell you exactly what's working and what to improve — lighting, expression, angle, and more.
Method

How does the AI rate faces?

On desktop, the attractiveness test loads a 68-landmark facial detection model into your browser via TensorFlow.js — derived from the iBUG 300-W landmark spec used in academic facial-analysis research (on mobile, the same landmarks are measured in memory by our scan server and the photo is deleted instantly).

The model locates:

  • the corners of your eyes
  • the nasal tip
  • the philtrum
  • the zygion (cheekbone) points
  • the gonion (jaw angle) points
  • the upper and lower lip border

From those landmarks, 17 measurement passes run.

Each metric (symmetry deviation, canthal tilt angle, FWHR ratio, midface thirds, jawline angle, philtrum-to-chin ratio, eye spacing, nasal width, lip-fullness ratio, etc.) is normalized by interpupillary distance — the same normalization used in clinical craniofacial papers — so the score is independent of how close you are to the camera.

Each metric score places your measurement against the first scans of about 2,100 RealSmile users, so 50 is the median scanner. The report scores every measurement your photo supports (typically 15 of the 17). There is no overall geometry score: averaged together, the measurements correlated at r ≈ −0.33 with the 1,566 benchmark faces' human ratings, no positive relationship, so that score was retired. The attractiveness estimate is a separate number: your Face Score, a percentile from a face-appearance model validated against averaged human ratings (see the validation).

For executive-presence use cases — LinkedIn primary, About-page lead, sales-deck cover — the question is how the photo reads for competence and approachability rather than for a dating swipe. Visitors optimizing for work, not dating, can jump to the LinkedIn / Executive Pro Audit ($49).

How this attractiveness test compares

Most attractiveness tests online fall into two buckets: question-based quizzes that never look at your face, and phone apps that require a download and an account before showing a score. We have not benchmarked those tools and make no claims about their accuracy — but here is what you can verify about this one in about 60 seconds:

What you can verifyHow
Free, in the browserNo app download, no signup, no account before you see a result.
17 metrics measured, not guessedSymmetry, canthal tilt, FWHR, jawline angle, and 13 more are computed from 68 facial landmarks on your actual photo — each one shown to you, not hidden behind a single number.
A validated percentileThe Face Score percentile is checked against averaged human ratings: r ≈ 0.80 on 1,566 benchmark faces, and r ≈ 0.40 on 46 real phone photos rated by our users, which is closer to what you will see (see the validation).
Private by designThe measurement runs in your browser on desktop; mobile photos are processed in memory and deleted immediately. Before scoring, your photo is checked automatically for anything that should not be scanned. That check happens on our server and the photo is discarded straight after it. To run that check, our server sends the photo to outside AI services (OpenAI, and Anthropic, which checks that a face is present and is OpenAI's backup) that only return a verdict on it; the check itself keeps nothing. By default nothing is kept or used for training.

If another test suits you better, use it — the fastest way to decide is to run this one, look at the 17 measurements it shows you, and judge whether the breakdown is useful.

How to read it

What does my attractiveness score mean?

ScoreWhat it means
0–40Check the photo first. Flat or harsh lighting, a tilted head or a low camera angle all distort the landmarks the scan reads. Retake with even light, the camera at eye level and a neutral expression before reading anything into the number.
40–60Average. A solid baseline. For the cleanest reading, use soft natural light, an eye-level camera, a level head and a relaxed, neutral expression.
60–80Above-average first impression. Good symmetry and proportions, with the photo conditions working in your favour. Keep this photo's light and angle for any rescan you want to compare.
80–100Top-decile photo. This shot is creating an excellent first impression. Use it as your dating-app lead, LinkedIn headshot, or professional bio photo.
How to read your score▾

Score bands describe geometry conformance (0-100, how close your measurements sit to their ideal ranges) — not a population rank, a percentile, or a clinical diagnosis. Each band names the dominant lever to pull next.

BandReadDominant lever
0–40Developing geometryRule out the photo first — flat overhead light, a tilted head, or a low-angle shot all distort the read. Retake in even light at eye level and compare the two before drawing conclusions.
41–60Mixed geometryA clean photo gives the most accurate read — softer lighting, eye-level camera, relaxed brows, upright posture. Expect a few points of movement between photos of the same face; the structure itself does not change.
61–80Strong geometryThe structure is already reading well. The remaining lever is photo polish — outfit and setting consistency, color grading restraint, and using a single lead shot across platforms.
81–100Very strong geometryMarginal returns from face-targeted changes. The leverage moves to situational presentation — context of the shot, what you're doing in it, and how the photo set tells a coherent story.

Geometry conformance (0-100), not a population rank.

How RealSmile compares to other attractiveness tests

Your photo is processed and deleted right away. Nothing is kept unless you choose to save it.

TestPrice · Speed · Privacy
RealSmileFree forever · InstantPrivate by default
attractivenesstest.comFree with ads · InstantPhotos uploaded to server
LooksMax AI App$9.99/week · InstantPhotos uploaded to server
Photofeeler$0.20/vote · Hours to daysPhotos shown to strangers

Prefer to compare yourself with friends rather than with a benchmark? The mog rating scores everyone in one group photo.

What people use the attractiveness test for

Use caseWhat the test tells you
Dating App PhotosFind out which photo makes the best first impression. Your lead photo determines your match rate on Tinder, Hinge, and Bumble. Analyze my face →
LinkedIn & ProfessionalChoose the headshot that signals confidence and approachability. Analyze my face →
Social MediaTest which selfie reads best before posting. Stop second-guessing and let data decide which photo to share. Analyze my face →
Self-ImprovementRescan in the same light and at the same angle to compare photos fairly and learn which setups show your face best — useful for building photo confidence. Start improving →

How to improve your attractiveness score

The scan reads your facial structure, and a photo can't change that — but light, angle and expression change how accurately a single photo shows it. Your Face Score typically moves about 11 points between two photos of the same face, so give the scan a clean read and compare more than one photo.

ChangeWhy it moves the score
Lips gently closedKeep a neutral expression with your lips gently closed. A smile or parted lips hides your lip proportions, so the scan leaves that measurement out rather than guess.
Use natural lightingSoft light from a window is universally flattering. It fills in shadows, makes skin look better, and creates warmth. Avoid overhead fluorescent lights.
Camera at eye levelPhotos taken from below exaggerate chin and nostrils. Hold your phone at eye level or slightly above for the most symmetrical, flattering result.
Relax your faceTension in your forehead, jaw, and around your eyes makes photos look stiff. Take a deep breath and loosen your jaw before shooting.

Attractiveness Test FAQ

How does the AI attractiveness test work?

The test runs our free face scan. It finds 68 facial landmarks, calculates 17 facial-geometry measurements (symmetry, canthal tilt, facial thirds, jaw taper and more), each read against a reference range, and gives you your Face Score — a percentile checked against averaged human ratings: r ≈ 0.80 on 1,566 benchmark faces and r ≈ 0.40 on 46 real phone photos rated by our users.

Is this attractiveness test really free?

Yes — the scan and your Face Score are free: 1 scan a day with no signup, 5 with a free account. The full breakdown of your measurements, with a PDF and a 30-day plan, is an optional $14.99 one-time report.

Do you upload or store my photos?

On desktop, the measurements run in your browser. On mobile, your photo is processed in memory by our scan server and deleted immediately after the landmarks are measured — never written to disk. Before scoring, the photo is also sent once to our server for an automated safety check; we don't keep it.

To run that check, our server sends the photo to outside AI services (OpenAI, and Anthropic, which checks that a face is present and is OpenAI's backup) that only return a verdict on it; the check itself keeps nothing.

By default nothing is kept: no photo, no copy, nothing used to train anything.

There are three exceptions and all three are your choice — you can save a scan photo to your own account, you can submit a photo to the community rating pool, and the paid Audit keeps a thumbnail so you can reopen your report.

Each one is deletable, and deleting removes the file itself, not just the link to it.

Can I really improve my attractiveness score?

Mostly, what changes between scans is the photo. The same face's Face Score moves a median of 11 points between two photos, and a third of people see 20 or more, because lighting, angle and expression change what a single photo shows. Compare rescans taken in the same conditions before reading anything into a change. The structural measurements — proportions, symmetry, jaw and eye geometry — do not change with practice.

How is this different from "am I pretty" tests online?

Most online attractiveness tests upload your photos to their servers and give vague feedback. RealSmile is private by default — your photo is processed and deleted unless you choose to save it — gives specific actionable tips, and gives you measured facial geometry with reference ranges and a percentile validated against human ratings.

Can I compare photos to find my most attractive one?

Yes. Upload 2-6 photos and the AI ranks them from most to least attractive, with specific scores and tips for each. This is perfect for choosing your lead photo for dating apps or LinkedIn.

How is this different from Photofeeler, QOVES, Aurale or Umax?▾
ToolPrice (source) · what you get
Photofeeler~$0.20 per vote (their 2026 pricing page)Crowdsources human votes. Hours-to-days latency. Useful for absolute attractiveness ratings from real people; not useful when you need a verdict before posting tonight.
QOVESYearly membership (qoves.com, 2026)Human-reviewed reports. The depth is real, but the membership gates first-time users.
Aurale$49 one-time, 30-day refund (testimonial copy indexed on aurale.app, 2026)Photo uploaded to their server.
Umax$3.99–$24.99 weekly/monthly subscriptions (Adapty paywall ladder, 2024 ledger)Photo uploaded to server.
RealSmile attractiveness testFree; optional one-time upgrades: a $14.99 report and $19.99 / $49 auditsInstant, and your photo is deleted after analysis unless you choose to save it. Compare RealSmile vs the QOVES alternative or pick your best photo with the free dating-photo ranker.

What you get free vs paid

TierWhat you get
Free forever
  • Face Score percentile
  • Your strongest metric
  • Up to 6 photos compared side-by-side
  • 1 free scan a day, no signup (5 with a free account)
  • Private by default — deleted after analysis unless you choose to save it
$19.99 / $49 one-time
  • $19.99 — Dating Photo Audit: up to 10 photos ranked by Face Score with each photo's strongest and weakest metrics + 5-page PDF
  • $49 — Pro Audit (LinkedIn / Executive): same engine, read for competence and trust, plus a reshoot brief

One-time. No subscription.

Privacy: nothing kept by default

On desktop, the attractiveness test loads its model into your browser via TensorFlow.js and runs the 17-metric analysis on your device. On mobile, your photo is processed in memory by our scan server and deleted immediately after the landmarks are measured — never written to disk. By default nothing is kept and nothing is used to train anything.

This is a structural privacy difference vs Photofeeler, QOVES, Aurale, and Umax (all four upload and retain your photo on their server, per each provider's 2026 privacy policy).

Attractiveness is the score. Looksmaxxing is the roadmap.

Find out exactly which of your 17 facial metrics are pulling your score up — and which are holding it back.

Jawline, canthal tilt, facial thirds, eye shape, FWHR — each metric scored 0-100 with specific improvement steps in the full report. Your Face Score percentile and strongest metric are free.

How attractive am I? Find out now.

Free, instant, private. Upload a photo and get your attractiveness score with actionable feedback.

Premium Dating Photo Audit · Delivered in 1–2 minutes

You read the page. Now turn the score into a plan.

Attractiveness test is the entry. The $19.99 Premium AI Dating Photo Audit is the deliverable: 17 metrics, lead pick, deletion flags, 5-page PDF in 2 minutes.

✓ 1–2 min delivery · ✓ 17 metrics measured · ✓ Lead-photo verdict · ✓ Stripe secured

Variant tests

The same 17-metric scan on every page, with a reading guide for a specific context. Pick the one that matches your situation.