The Best Ways and Apps to Track Body Composition in 2026

The short answer: the most accurate way to measure body composition is a DEXA scan, but the most useful way to track it is whatever you’ll repeat under identical conditions every week. For most people that means pairing a consistent home method with an occasional reference point:

• a photo-based app,
• a smart scale used at the same time each morning,
• or a tape-measure equation.

No app or scale matches DEXA for a single reading, so compare each method only with itself.

How this comparison was put together

• Methods: described from the research literature, cited below.
• Apps: included if they estimate body composition from photos on a consumer phone app and have a public product website. Details are limited to what each company publishes about itself, checked September 2026. Features, pricing and claims change; check the company’s site before deciding.
• Accuracy claims: only independent research is presented as evidence. Company claims are labeled as company claims.

Part 1: the methods

Method	What it measures	Evidence on accuracy	Repeatable at home?	Best for
DEXA scan	Fat, lean soft tissue, bone, by region	Common reference method; results differ between machines and software	No: clinic visit	An occasional anchor, 1–2× a year
Smart scale (BIA)	Estimates fat from electrical resistance	Weight accurate; fat mass underestimated by a median of 2.2–4.4 kg vs DEXA (2021). Highly repeatable, but single-reading error −3.5 to +11.7 pp across 15 devices (2023)	Yes	Daily weight; body-fat direction under identical conditions
Skinfold calipers	Subcutaneous fat at set sites, via equations	Depends heavily on measurer skill and consistency	With a trained, consistent measurer	Gym assessments by the same person
Tape-measure equation (Navy)	Estimates body fat from circumferences	Population equation; several points of individual error	Yes	A free baseline and waist trend
Photo-based app	Estimates composition from visual markers	Good group-level accuracy, weaker individual accuracy, limited long-term tracking evidence (2026 review)	Yes	Weekly trend plus a visual record

If you want the free option first, the body fat calculator runs the Navy tape-measure and Jackson-Pollock skinfold equations. DEXA vs. calipers vs. BIA vs. photo goes deeper on each method.

Part 2: photo-based apps

App	What it says it does (from its own site)	Platform	Pricing (as published)	Accuracy claims (company’s own, not independently verified)
Corpus (ours)	Guided photo check-ins with capture-consistency checks; estimated body fat, FFMI, muscle-group views, composite score; AI Coach; optional Apple Health recovery context; no ads	iPhone (iOS 26+)	Subscription: monthly or yearly, with a trial on yearly	No published validation yet. An internal repeatability study is in progress and will be published with its sample size.
Bodilab AI	Single-photo estimates of body fat, lean mass, FFMI, BMI, waist-to-hip, 12 muscle groups; trend charts; AI coaching	iPhone	Free tier; Pro subscription with a 3-day trial	Says it’s best at tracking change rather than absolute values; describes results as estimates
Spren	Selfie-based body fat, lean and fat mass, fat distribution, resting metabolic rate; nutrition targets; AI coaching	Phone app	Not disclosed on its homepage at time of checking	Says it uses validated algorithms and cites peer-reviewed publications

Where Corpus is different

• Capture consistency is built into the check-in. Photo Match checks framing, pose, distance and lighting against your own baseline and flags check-ins that differ. That matters because, as the research review found, individual-level accuracy is the weak point of photo estimates. A consistent capture is what makes a trend readable.
• Composition plus recovery. Optional Apple Health signals (sleep, HRV, resting heart rate) sit next to the composition trend.
• Stated limits. Corpus doesn’t claim DEXA-level accuracy, and the methodology page says what the estimates can and can’t do.

What to know about photo-based estimates in general

A 2026 review in Nutrition identified eleven research-backed body-composition apps. It found good accuracy for body fat percentage at the group level, but:

• reduced accuracy for individuals and in people with higher body fat
• insufficient validation across diverse populations
• limited evidence on tracking change over time

A 2025 preprint on one smartphone computer-vision app, not yet peer reviewed, summarized its finding in the title: reliable but biased compared with BOD POD and InBody. That’s the same pattern seen in smart scales.

In practice: whichever app you choose, use it the same way every time and read the trend, not the number.

Which should you use?

• You want the most accurate single number: DEXA, once or twice a year, same machine.
• You want free: a tape measure, the body fat calculator, a weekly waist measurement and consistent photos.
• You already own a smart scale: use it at the same time every morning and trust the direction, not the level.
• You want a weekly composition trend plus a visual record: a photo-based app. Try the free tiers if cost matters, and pick the one whose capture routine you’ll actually repeat.
• You’re on iPhone and want composition, capture consistency and recovery in one place: that’s what we built Corpus for.