FaceIQ Labs review: what it costs, what the terms actually say, and the five different products called FaceIQ

FaceIQ Labs is one of the most searched names in facial analysis, and almost nothing written about it contains a price. That is not an oversight by the writers — the price is genuinely not on the sales page. It is, however, written down. This review is the result of reading the whole document it is written in.

Before anything else, work out which one you are looking at, because the search results mix them and they have different owners, prices and business models. Someone complaining about billing on Reddit and someone praising a $50 report on Gumroad are not talking about the same company.

The rest of this page is about the first row only.

There is no pricing page. `faceiqlabs.com/pricing` returns a page with almost no text on it, the homepage says only 'Try Free', and the price is not rendered by JavaScript either — it simply is not on the marketing site. This is why every third-party write-up either omits the price or estimates it, and why one competitor's comparison page lists '~$120 annually (approximate, per reviews)'.

The real figures are in the Terms of Service, in the Partner Program section, where the referral reward is defined by naming each plan and its monthly rate: Basic plan partners ($24.99/month) and Pro plan partners ($49.99/month). Those are the two tiers.

The number that matters is the one on the right. A monthly price of $24.99 sounds like a small decision; $299.88 is a different kind of decision, and $599.88 for Pro is a different one again. Because this is a self-improvement programme measured in months — the homepage's own headline case runs '2.5 years of real commitment' — the twelve-month column is the honest way to read the price, not the one-month column.

This is the single most surprising thing in the document, and it is stated plainly rather than hidden. Free trials, the terms say, 'are NOT generally available to all users'. They are 'exclusively available as a reward for partners who successfully refer a paying customer'. And then, in case that leaves room: 'There is no other way to obtain a free trial — users cannot sign up for a free trial directly.'

So the path to a trial is: join the Partner Program, persuade someone else to buy a subscription, wait for their first payment to clear, and receive a trial of 'typically 3 days'. To claim it you must 'select a subscription plan and provide valid payment information', and at the end it 'will automatically convert to a paid subscription at the then-current rate unless you cancel'.

On the surface this is a standard 14-day refund policy. The condition attached to it is what changes its meaning:

“We offer a 14-day refund period for new monthly subscriptions, provided that you have not used the Service. 'Use of the Service' is defined as generating at least one facial analysis. If you have not generated any analyses within 14 days of your purchase, you are eligible for a full refund. After 14 days, or once you have generated your first analysis, all payments are final and non-refundable.— FaceIQ Labs, Terms of Service, section 4

Generating a facial analysis is the only reason anyone subscribes. So in practice the refund is available to people who paid and then never used the product — and to nobody else. The same condition governs quarterly and yearly plans, where the sums are much larger and 'no partial or prorated refunds are available for unused portions'.

This clause is also the most likely explanation for the shape of the public feedback, which is worth looking at directly.

Yes, in the sense the question is usually asked. There is a registered company presence, a published Terms of Service, an app-store listing, a LinkedIn page and a Trustpilot profile. Nobody is taking your money and vanishing.

The reviews are a separate question, and they are unusual:

A distribution like this — nearly half at the top and nearly half at the bottom, with almost nothing in the middle — usually means two different experiences rather than one mediocre one. Read the negative reviews and they are not about the analysis being wrong. They are about money: one describes being charged after cancelling weeks earlier, another describes a charge arriving after cancellation, a third says customer service did not reply. Eighteen reviews is a small sample and you should weight it accordingly, but the complaints point at the same mechanism the terms describe, which makes them more credible than an isolated grievance would be.

The practical protection: if you subscribe, cancel through the account settings rather than by email, screenshot the confirmation, and check your card statement on the next renewal date.

Taking the product on its own terms, the offer is not a measurement — it is a system built around a measurement. The homepage organises 100+ metrics into four named pillars and layers tooling on top.

Split that list in two and the value question answers itself. The ratios and the score are arithmetic on landmark coordinates: deterministic, reproducible, and free in a dozen places. The roadmap, the tracking history, the chat and the community are hosting, product work and moderation. If you are paying $299.88 a year, you are paying for the second list. Whether that is good value depends entirely on whether you will use it for a year, and the homepage's own 30-month case study is an admission that shorter than that is not the intended experience.

Your photo is uploaded to FaceIQ Labs, so what happens to it is a matter of its terms. Those terms include a written biometric retention and destruction schedule — the specific artefact the Illinois Biometric Information Privacy Act requires — alongside a broad licence over what you upload.

The structural point is that all of this is necessary because the photo leaves your device. A written destruction schedule is a promise about a file somebody else is holding. Running the landmark detection in the browser, so the image never leaves the machine, removes the need for the promise. That is how the tools on The Face Analysis work: the photo is not uploaded and no score is saved on our servers, so there is no file to retain.

Two very different claims get bundled into one number, and separating them is the most useful thing you can do before paying for any tool in this category, including ours.

The measurements are real, and they are fragile

Landmark positions and the ratios computed from them are arithmetic. Given the same photo, any competent implementation returns the same numbers. What changes them is the photo: a face turned 15 degrees off-axis projects differently onto the sensor, and every horizontal ratio shifts with it. Lens focal length changes apparent nose projection. Overhead lighting deepens every shadow the score reads as a fold. This is why 'my score changed' posts are so common, and it is rarely the algorithm.

It also sets a floor on how much precision is worth chasing. Research on facial asymmetry puts the threshold at which observers begin to notice a deviation at roughly 3 mm — larger than the differences most tools report to two decimal places.

The score is a choice, not a finding

A single number out of 100 is a weighted sum, and somebody picked the weights. There is genuine research behind some of the inputs: averageness and symmetry both show reliable associations with rated attractiveness across studies, and skin quality is one of the strongest single predictors in the literature. But no published work validates any particular commercial weighting as a predictor of how people actually respond to a specific face, and the golden ratio in particular — the most marketed idea in this field — does not survive scrutiny as a beauty criterion.

So a score is useful as a repeatable index of your own face against itself over time, which is exactly what progress tracking is for. It is not a measurement of how attractive you are, and no amount of decimal places makes it one.

The honest split: everything geometric is free somewhere, and the programme is not.

• ✓Price. Every measurement is free and unlimited, with no account. FaceIQ Labs is $24.99 or $49.99 a month — $299.88 to $599.88 a year at the monthly rate. Paid plans here start at $7.99 a month, or one-off credit packs from $4.99 with no subscription.
• ✓You can try it without paying. The measurements are free to run as often as you like. FaceIQ Labs' free trial is available only to partners who refer a paying customer.
• ✓A refund you can actually use. Unused credits are refundable with no time limit. FaceIQ Labs' refund ends the moment you generate your first analysis.
• ✓Your photo stays with you. The landmark model runs in your browser, the photo is not uploaded and no score is saved on our servers — so there is no licence over your upload and no three-year retention window.
• ✓Every score has a reference. The attractiveness score, symmetry, fWHR and midface ratio are placed against 102 adults from the Face Research Lab London Set, measured with the same code; the distributions are published.
• ✓The formula is public, and so is its test. 478 landmarks, symmetry weighted 40% and three ratios 20% each, written out on the methodology page. On those 102 faces the composite did not predict 2,513 people's ratings (ρ = −0.07), and the page says so.
• ✓It names what a photo cannot measure. Each tool page says which traits need a profile view or depth, instead of guessing.

One last note on the search results that brought you here. Almost every page ranking for this brand is either the company's own property or a platform where anyone can post. There is no independent editorial coverage, which is why the price was still unknown to the internet on the day this was written even though it has been sitting in the terms all along. Read the terms of anything you are about to subscribe to — in this case it was the only place the number existed.