Support
FAQ
Last updated: May 2026
How accurate is photo-based body fat measurement?
Reliable for tracking your own progress over time, less so for absolute clinical accuracy. LeanBody compares your photographs against patterns recognised by a frontier vision model and scores against the NHANES 2017-20 DXA reference cohort. Treat the absolute number as approximate; trust the trend across multiple scans. Not DEXA-grade, not a medical device.
What photographs do I need to take?
Four photographs: front flexed, front relaxed, back flexed, back relaxed. Take them shirtless or in a sports bra, with shorts, on a plain background, with the camera at chest height capturing the full body head to feet. Soft, even lighting from the front. The whole sequence takes about 90 seconds on a phone.
Do you store my photographs?
No. Photographs are uploaded to ephemeral blob storage so the vision model can analyse them, then deleted at the moment results are delivered. We have no persistent database. Your numerical results live only in your browser's localStorage and auto-delete after 48 hours. We never see the photos and cannot retrieve them.
How does LeanBody compare to a DEXA scan?
DEXA scans are the radiologic gold standard at $80-150 per appointment and roughly 30 minutes in a clinic. LeanBody runs in 90 seconds on your phone for $5. The trade-off is precision: DEXA is more accurate at a single point in time; LeanBody is far cheaper for tracking trend across multiple scans.
Is LeanBody more accurate than a smart-scale measurement?
Generally yes for tracking trend. Smart scales use bioelectrical impedance (BIA), which depends heavily on hydration, food intake, glycogen stores, and time of day — readings can vary several percentage points within a single morning. Photo-based estimation is not sensitive to those variables, so the trend is more interpretable across a cut or bulk.
How does the AI estimate body fat from a photograph?
A multimodal vision model evaluates all four photographs simultaneously, considering visible muscle definition, vascularity, subcutaneous fat distribution, and bone landmarks. Your self-reported height and weight calibrate the result against the NHANES reference cohort. The model is not measuring tissue — it recognises visual patterns associated with body composition ranges.
What is FFMI and how is it calculated?
Fat-free mass index (FFMI) is your lean body mass in kilograms divided by your height in metres squared. It was proposed by Kouri, Pope, Katz and Oliva (1995) as a way to quantify muscularity independent of body fat. Typical adult ranges: 18 (untrained) to 23 (well-trained); above 25 is considered exceptional or non-natural.
What is the natural FFMI ceiling?
Around 25 for men and 22 for women, per the original Kouri et al. 1995 study comparing natural lifters with enhanced bodybuilders. FFMI values above this are achievable with anabolic-steroid use; without enhancement they are extremely rare. Most well-trained natural lifters with years of dedicated work plateau in the 22-24 range.
What's a good body fat percentage for men?
For men, conventional bands are: 6-9% stage-ready, 10-13% cut, 14-17% athletic, 18-22% average, 23-27% soft, 28% and over heavier. The U.S. NHANES 2017-20 male median sits around 25%, with an interquartile range of roughly 21-31% — the middle 50% of adult men land between those values.
What's a good body fat percentage for women?
For women, conventional bands are: 14-17% stage-ready, 18-21% cut, 22-25% athletic, 26-30% average, 31-35% soft, 36% and over heavier. The U.S. NHANES 2017-20 female median sits around 36%, with an interquartile range of roughly 32-41% — the middle 50% of adult women land between those values.
How long do my results stay accessible?
48 hours after a scan, your results are auto-removed from your browser's localStorage. There is no server-side copy. To keep a permanent record, save the shareable plate as an image or screenshot the results page during that window. We do not provide account-based history because there are no accounts.
Why $5?
$5 covers the cost of running a frontier multimodal vision model with reasoning, plus Stripe processing fees, with a small margin. There is no subscription, no account creation, no upsell. The price is deliberately set at the level of a coffee so a scan is a casual decision, not a commitment.
What if I'm not satisfied?
Email hello@leanbody.io and request a refund. No questions asked. We are not interested in coercing you to keep a result you do not trust. The $5 charge will be reversed via Stripe and there is no follow-up sequence — we do not have your email beyond the receipt anyway.
Is LeanBody a medical device?
No. LeanBody is not a medical device, not FDA-cleared, and not intended for diagnostic, therapeutic, or clinical use. The output is informational only. Do not use it to make decisions about disease, eating disorders, or any clinical condition. Consult a licensed practitioner for medical guidance about body composition.