Potential AI App Review: Is It Worth It? An Honest Look at 'Look a 10'

You typed potential ai and got generic AI pages. This covers Face Rating: Potential AI ("Potential AI" on Google Play), the looksmaxxing app whose store image shows a man above a seaside town, a 10 in every column: See your 10/10 potential.

Short answer: a real app on normal Apple and Google billing, worth it only on the $19.99–$24.99 yearly plan, as a routine tracker. Its "Look a 10" image is an AI edit of your photo, not a forecast of your face.

Disclosure: we sell a first-impression test (full report $5) and haven't bought this app's plan. Facts checked October 17, 2026.

Potential AI at a glance

• Price: Pro $4.99 a week to $24.99 a year; $2.99 scans; $4.99 "Look a 10" (App Store).
• Paywall: at the scan, reviewers say.
• Rating: 4.8 from 40 iOS ratings; 4.4 from about 1.8K Google Play reviews.
• Privacy: photos go to Google Gemini.
• Verdict: cheap yearly habit tracker; its terms call it entertainment.

What is Face Rating: Potential AI?

A looksmaxxing (maximizing your looks) app: upload a selfie and side profile, get scores, follow a 30-day plan. It reached Google Play on January 27, 2026 and the App Store (seller: Muhammad Younas Rehmatullah) on February 13.

• Scores: overall, jawline, cheekbones, skin, masculinity, plus a "PSL Rating" from the forum PSL scale; one screenshot shows 6.50/8.0, "Chadlite."
• Routine: skincare, grooming, mewing (tongue-posture) drills and a daily face upload with a streak counter.

Is Potential AI free? Where does the paywall hit?

Free to download, not free to use in full. The developer's pricing page promises "1 AI Face Scan per month" at $0, but Google Play reviewers describe paying for the scan:

1. Feb. 23, 2026 (5★): "The first app that rates you without asking for subscription."
2. Apr. 8 (1★): "scanning is behind the paywall."
3. Aug. 5 (2★): "you have to pay for everything and not even 1 thing is free."

The earliest reviewer got a free rating; later ones hit the price before the scan (why apps do it).

We didn't install or pay; your version may differ.

How much does Potential AI cost over a year?

$19.99–$24.99 on an annual plan, $119.88 on monthly, or $259.48 if the $4.99 weekly plan renews for 52 weeks.

Credit where due: $4.99 a week is the median weekly price in our pricing index of 15 apps, and $24.99 a year for a habit you keep is modest.

It buys a routine, not a verdict on your face.

What is the "Look a 10" filter, really?

A generative-AI edit. The privacy policy says photos are "sent to Google's AI service for analysis and image generation" and calls the output "AI-enhanced versions of your photos." The website promises a "visualization of your maximum potential based on your unique bone structure and genetics."

Our name for the gap: a render is not a forecast. A forecast would need data on how real faces change with fat loss, skin care and age. Generative AI instead produces a plausible image of what its training data treats as attractive, seeded by your face. It can show you a haircut. It can't read your genes.

Two tells. The store screenshot of "Your 10/10 Look" scores the render 10 in every column: the system grades its own edit. And the example is styled (linen shirt, soft light), which matters: attractiveness ratings vary more across photos of one person than across different people (Jenkins et al., 2011).

Concession: the one written iOS review (5★, Feb. 13, 2026) called the filter "insane." As a haircut mood board, fine; as a target, unbeatable by design.

What do users say about Potential AI?

Short praise, plus a recurring complaint about paying to scan. Forty iOS ratings and one written review prove nothing either way. Of the 160 recent and most-relevant Google Play reviews we pulled on October 17, 122 ran three words or fewer.

• Best case: "it actually helped me make me a bit better in 90 days but three months won't make you perfect" (5★, Aug. 17, 2026).
• Scores: "it gave me 0 for everything" (1★, Apr. 2, 2026).

Is the Potential AI score accurate, and accurate at what?

Unknown: no validation or repeat-test data is published (we didn't run a repeat test either), and the documents disagree on what does the scoring. The features page credits a "proprietary computer vision model"; the privacy policy names Google Gemini. Google Play promises "objective scores"; the terms say "We do not guarantee the accuracy of AI-generated analysis."

Concession: five free AI rating sites tracked 24 human raters at r = 0.84 on 40 AI-generated female faces, but scored higher, 6.9 vs. 5.0 (Goshtasbi et al., 2024). No men, no real selfies, not this app. And strangers never see your selfie; they see you, moving.

Is Potential AI safe? What happens to your photos?

Billing is ordinary, and the policy (updated Feb. 5, 2026) is clearer than most: photos go to Gemini, which, the developer says, doesn't retain them; saved scans stay until you delete your account; no AI training "without explicit consent."

• Labels disagree: the iOS label lists no photos; Google Play's lists "Photos and videos." Both are self-declared (what labels miss).
• Age: 9+ on iOS, "Everyone" on Google Play; the terms say 13+.
• Deletion: Profile → Settings → Delete Account (details).

How do you cancel Potential AI and get a refund?

Through Apple or Google (full guide). Deleting your account "does not automatically cancel your subscription," the developer warns; neither does uninstalling.

1. iPhone, cancel: Settings → your name → Subscriptions → Cancel Subscription (Apple).
2. iPhone, refund: reportaproblem.apple.com → "I'd like to" → "Request a refund"; expect an update in 24–48 hours (Apple).
3. Android, cancel: Google Play → Payments & subscriptions → Subscriptions (Google).
4. Android, refund: "Report a problem" in your order history within 48 hours; after that, ask the developer (Google).

Who should use Potential AI, and what should you use instead?

Use it for a cheap yearly routine, if you hold the numbers lightly. Skip it to learn how you come across; no score or render measures that.

• No paywall: our list of looksmaxxing apps with no paywall.
• The missing axis: Potential AI grades and redraws a face. Our test reads your whole first impression, face, physique and outfit. The analysis runs before you pay and shows your band, appeal personality and one strength. The full report (your read and natural ceiling on a 70–155 axis, how many of 100 women passing would notice you, a dated 7/30/90-day plan) is $5 once, no subscription. It isn't a validated clinical instrument either.

A daily face upload with a streak counter can turn self-care into self-surveillance. If checking leaves you feeling worse, step back (warning signs).

Key numbers

• 5 weeks of $4.99 billing ($24.95) cost about as much as the yearly plan ($24.99).
• 1 written review behind 40 U.S. App Store ratings.
• 122 of 160 Google Play reviews we pulled ran three words or fewer.
• 100 ms: enough of a face for judgments that track unhurried ones (Willis & Todorov, 2006).

The bottom line

Potential AI is real, cheap on the yearly plan and open about its AI provider. The weak link is the promise: a Gemini edit sold as your "maximum potential." A render is not a forecast.

Your face doesn't have a score that decides your life. It has an effect on people, formed in a fraction of a second and more movable than any number admits. See where your read lands; the analysis runs first, and the full report is $5, once.

Sources

Checked October 17, 2026: App Store, reviews, Google Play, and Potential AI's privacy, terms, pricing and features.

• Goshtasbi, K., Hakimi, A. A., & Wong, B. J. F. (2024). Artificial intelligence versus human focus group rating of facial attractiveness. Facial Plastic Surgery & Aesthetic Medicine, 26(4), 371–376. doi:10.1089/fpsam.2023.0281
• Jenkins, R., White, D., Van Montfort, X., & Burton, A. M. (2011). Variability in photos of the same face. Cognition, 121(3), 313–323. doi:10.1016/j.cognition.2011.08.001
• Willis, J., & Todorov, A. (2006). First impressions: Making up your mind after a 100-ms exposure to a face. Psychological Science, 17(7), 592–598. doi:10.1111/j.1467-9280.2006.01750.x