Best free looksmaxxing app with no paywall (2026) — Mogged vs Umax vs LooksMax AI, compared

TL;DR

• The answer: none of Mogged, Umax or LooksMax AI shows you your score for free. All three are free to scan and paid to see, and the wall lands after your photo is up — that's the business model, not a bug.
• Where free actually exists: outside the app stores — a browser test, a human-feedback site like Photofeeler, or a general AI chatbot with the right prompt. All three are below.
• If it must be an App Store app: LooksMax AI is the only one where the score unlocks without money — you pay in contacts, by inviting friends.
• Ours, for completeness: browser-based, no install; score, full read and your biggest levers free. Our own wall is a free account — no card — and, importantly, the analysis is finished and waiting before you're asked for an email. Going deeper is about $1.50 a module (smallest top-up $5), no subscription.
• Skip all of it if the number is making you feel worse rather than clearer.

All prices, download counts and paywall placements below were last reviewed in September 2026, against the public App Store listings and product pages. This is desk research: we read the listings and user reviews, we have not paid for any of these subscriptions. Check the price sheet on your own screen before tapping.

You typed "best free looksmaxxing app no paywall 2026" for a reason. Probably this one: you tried one, watched the scan crawl to 94%, and a screen slid up asking for $3.99 a week — before the number you came for.

Already been charged? Do that first, before you read anything else — a weekly subscription renews while you research. On iOS it is Settings → your name → Subscriptions, and refunds go through reportaproblem.apple.com; on Android it is Play Store → profile → Payments & subscriptions. The step-by-step, including what to write in the refund box, is in how to cancel a Umax or LooksMax subscription.

Full disclosure: we build the free first-impression test linked at the end — read this comparison knowing that.

The comparison, in one table

Photo handling is deliberately not a column here — it needs more than a table cell, and every one of these rows uploads your face to somebody's server, ours included. Where your photo goes further down is the honest version, including the one tool on this site that never uploads anything. One app in this category makes the paywall question unusually hard to answer, for reasons we set out in the JawMax app review.

The one-line verdict: none of Mogged, Umax or LooksMax AI will show you your score for free. If seeing the number without paying is the whole point, you have to leave the app stores.

If that's you and you want the shortest route out of the app stores: our browser test is the one we build — four quick picks, then your photos, and your score and read are free (a free account, no card). The non-ours options are three paragraphs down, and we'd rather you took one of those than paid $3.99 a week for a lock screen.

The four apps, in a bit more detail

Described neutrally — what they do, how they charge, what users report. Where a criticism appears, it is a quoted user or reviewer, not our verdict.

Umax. The category leader. Blake Anderson's app reportedly crossed 7 million downloads and scores your face 0–100 across jawline, cheekbones and "masculinity." The honest answer to is Umax free is "free to scan, paid to see." The loudest complaints are about consistency — users reporting they "submitted the same picture 3 times and almost every time a different number" and calling results "completely inaccurate". Psychologists quoted by Fortune have warned that handing teenagers a low face number may feed the youth mental-health crisis.

LooksMax AI. Often cited around 11 million downloads, viral largely through an invite-friends-to-see-your-result loop — which is a paywall paid in social capital rather than dollars. It maps landmarks (canthal tilt, gonial angle, "facial harmony") and outputs a score plus a fix list. Read its most recent App Store reviews before subscribing, and keep the cancellation steps handy if you do.

Mogged. It is in this page's title because people search it by name, so it deserves more than a shrug. Mogged is one of the long tail of near-identical scanners: same upload, same PSL-style number, same scan animation, same paid tier after the result is computed. What makes it worth naming separately is the framing — the brand leans on the forum vocabulary of being "mogged" by better-looking men, which sets the tone of the output as a ranking against other people rather than a read of you. On the numbers themselves it inherits the same instability as the rest of the category.

Mogged's price is the one figure we won't print. When we re-checked in September 2026, the site we had previously cited for it — mogged.link — no longer sells Mogged; it now serves a different product, Handsome AI. We couldn't find a Mogged price page we're willing to quote, so check the subscription sheet on your own screen before you tap.

QOVES Studio. A different shape: not a 99-cent app but a paid report service positioned as clinical-grade facial analysis — the benchmark the cheaper apps imitate (official site, Trustpilot reviews). Pushback here is about value, not price alone — the recurring theme in public reviews is that the report reads generic for what it costs, which you can read firsthand on Trustpilot rather than take from us. Not free, and not trying to be.

The actually-free routes (including ones that are not ours)

If your requirement is see something useful, pay nothing, ever, there are four honest options, and each has a real limitation. (For browser-based face-analysis tools in general — where their paywalls land and what to ask before uploading — see free face analysis.)

• A general AI chatbot. Upload a photo to ChatGPT, Claude or Gemini and ask for a plain read: grooming, lighting, hairstyle, what the photo communicates. Free tiers handle this. Limitation: these models are trained to be agreeable, they refuse or hedge on ranking faces, and you get no consistency between sessions. Good for actionable notes, useless as a score.
• Photofeeler. Real humans rate your photo on smart, trustworthy and attractive, and you earn credits by rating others rather than paying. Limitation: it costs you time instead of money, the ratings are of a photo rather than of you, and the sample is other Photofeeler users. Still, it is the only mainstream free route where actual human perception is in the loop. We also keep a list of Photofeeler alternatives.
• A free web face rater, such as Overchat's Rate My Face. Its tool page is titled "Free AI Face Rating" and returns a score out of 10 with a feature breakdown. Limitation: it scores photo geometry, not how you come across, and nobody has published a check of its number against real people.
• A browser test, ours included. No install, no store subscription. Limitation: ours is one, we built it, and no photo test is a validated clinical instrument — read the disclosure above and weigh it.

The chatbot prompt, since it's the one route above that depends on wording. Asked to "rate my face," most chatbots refuse or hand you a random number. Asked for a read, they're occasionally useful. Copy this:

Look at this photo the way a stranger would for one second.
1) First impression, in one sentence.
2) Three things helping the read (lighting, angle, grooming, expression).
3) Three things hurting it — with a concrete fix for each.
No number scores — one photo can't support one.

Expect a different answer in a new chat with the same photo; treat it as notes on your photo, not a measurement of your face.

Why the score keeps changing

This complaint runs through every app above, not one of them, and it is not a bug any of them can patch.

A looksmaxxing app measures the geometry of one still photo. Canthal tilt, jaw width, gonial angle — all wildly sensitive to things that have nothing to do with your face: lens focal length, how high you held the phone, side light versus overhead, mid-blink, how much you smiled. Tilt your chin two degrees and the jaw geometry shifts. The app is not lying on the second upload; it is measuring a different image.

Underneath that sits the bigger problem. The usual line is "nobody has ever checked whether these scores match real people." Somebody has, once, and the finding is worth knowing.

A team at UC Irvine scored 40 faces with five free AI rating sites and with a focus group of 24 otolaryngology residents. The two tracked each other closely (Pearson's r = 0.84, p < 0.01) — but the websites rated higher than the humans, averaging 6.9 against 5.0, with 30 of the 40 images significantly over-scored by the machines. The authors' own conclusion was to use these tools cautiously (Goshtasbi, Hakimi & Wong, 2024, Facial Plastic Surgery & Aesthetic Medicine).

Notice which way that cuts. If a scanner gave you a number that ruined your week, the only measured bias in the literature says the machine was probably being generous, not harsh. And notice how narrow the study is before you lean on it: 40 AI-generated, white, female, expressionless, front-facing images, judged by surgical residents — none of the apps on this page were among the five sites tested, and the comparison was between two groups looking at the same photograph.

Which leaves the claim that actually matters, and it survives intact: no one has ever checked any of these scores against how real people react to you in person. The models were trained to predict a "harmony" or PSL rating — a reference template built from a narrow slice of faces, as that study's own all-white, all-female sample quietly illustrates. So even a perfectly stable score answers the wrong question: how closely your bone structure matches a reference, not how someone reads you in the first second of meeting you.

Two things this does not mean. Geometry is not nothing — symmetry and structure carry some signal. And "it is all confidence and personality" is the opposite lie, which we are not selling. Looks matter; they matter as a perceived signal in context, not as a ratio scraped from one image. First impressions behave like a threshold rather than a ladder: past "well-kept, healthy, carries himself fine," extra decimal points stop deciding outcomes. We go deeper in why a high PSL face can feel like a 5 on a real date, in what women actually find attractive, and in whether looksmaxxing is pseudoscience.

Where does your photo go?

The listicles leave this off, and it is the part you cannot undo. You are uploading your face — not a password you can change, but your biometric likeness, often the same frontal portrait you use everywhere else online.

None of the listings above spell out photo retention in a way we can quote. So the defensive default: assume any upload leaves your device and may persist on a server, and read the App Store privacy label before you scan. The label tells you whether the developer collects photos and whether they are linked to your identity — it is two taps down the listing page and it is the only claim the developer is on the hook for. "We process locally" and "we do not store your photo" in marketing copy are claims you cannot audit.

The cleaner standard is analysis that runs in your browser so the image never leaves your device. That is rare, and we should apply it to ourselves before we apply it to anyone else: our main test does not meet it. It runs in a browser rather than an app, which spares you an install and a store subscription — but your photo is uploaded and analysed on our servers, and "in the browser" is not a claim we get to make about it. The one thing here that genuinely never leaves your device is our free canthal-tilt test, which is entirely client-side because it only measures your own facial proportions. If not uploading at all is your hard requirement, use that one and skip everything else on this page — what we do with your photo is the document to read before you decide.

Our own free-versus-paid line, stated plainly

You have been burned by a lock screen, so you are right to be suspicious of anyone in this space saying "free." Here is our boundary with no asterisk:

• Free: your score on our 70–155 perception axis (our calibration puts the typical man around 110 — it is not a mark out of 155), your face, physique and outfit read, your strengths, and which weaknesses are costing you the most. No card, no trial, no lock screen between you and the number.
• One thing that is a wall, and we'd rather say it here: opening your finished result needs a free account — an email and a code, no password, no card. Your analysis is generated and waiting before you're asked. That's a registration gate, not a paywall, and if you count it as a wall anyway, that's a fair way to count.
• What happens to that email afterwards, since it's the actual question. It gets used to sign you in and to keep your reports attached to you. There is no newsletter, no drip sequence and no promotional mail — our privacy policy lists what we use your information for, and marketing isn't on the list. The code you receive to log in is, in practice, the only mail you'll get from us. You can delete the account, and the photos go with it.
• Paid, and there is exactly one route. Individual deep-read modules open with credits at 150 each; credits are $1 per 100, $5 minimum, and they never expire — so a module runs about $1.50, and unspent credits stay yours. There is no timer, no bundle upsell, no subscription, no weekly billing, and nothing auto-renews. You can also spend them on a re-analysis (100 credits) or a question to the assistant (100), and nothing else on this site costs money.
• And yes, we have a refer-a-friend bonus too — 200 credits to each of you. We put LooksMax AI's invite loop in the table above as a paywall paid in contacts, so we owe you the distinction rather than the silence: their invites buy the score itself; ours only ever buy the extras. Your number is never held hostage to your address book. Judge whether that line is meaningful; we think it is the whole line.
• No App Store app. It's a web page: nothing to install, nothing attached to an Apple ID you'll forget, and no weekly billing machine to feed — a large part of why the score can be free. If you wanted a home-screen icon, we don't have one.

And what it measures is different from everything in the table: not how closely your bone structure matches a template, but how you are likely perceived in the first second — including the levers a static score throws away, like lighting, angle, expression and grooming. If you have already run a Umax number, how that score compares to real life is the natural next read.

Take the free test — score, full read and your top levers free with a free account. Going deeper is single modules at 150 credits, about $1.50 each. No subscription.

The checklist, if you are still going to try one

• No paywall on the score itself. This is the line that matters, and it is narrower than "no paywall anywhere" — most tools here, ours included, sell something. The test is whether the thing you uploaded your face for is behind the wall. An app that takes your image and then asks for $3.99 before showing you the number has the incentive backwards.
• A readable privacy label. If retention is vague, treat that as your answer.
• Perception, not just geometry. Useful output is "here is how you are likely read, and the levers that move it," not "your canthal tilt is positive, pay to see more."
• No verdicts handed to teenagers. A good tool gives a read and a direction, not a 16-year-old the news that his face is a 4.

If face-related anxiety is running your day rather than just nagging at you, no app — ours included — is the right fix. A real person is. A number from one selfie is a measurement of one photo under one light, against one narrow template. It is not a ceiling, not a sentence, and not how anyone who meets you will actually experience you.