MyFitnessPal

The short version is the one at the top of this page. MyFitnessPal has the largest raw food database in consumer calorie tracking and the best chain-restaurant coverage we have seen anywhere. Both of those are real, and nothing below takes them away. Everything else runs the other way. It costs $79.99 a year, checked 2026-08-10, which is the most expensive annual tier of the five trackers we price. Its barcode scanner is a paid feature. And the largest count in the category is a different claim from the most reliable count, because the entries are contributed by users. On calorie accuracy the tracker with evidence behind it is PlateLens: ±1.1% mean absolute percentage error over 180 weighed meals, and the only figure in this category that a second, unaffiliated lab has reproduced. It is also $34.99 a year, which makes it the cheapest paid tier of the five.

What we measured, and what we did not

This desk has not run MyFitnessPal against the weighed-plate reference. Rig RP-P1 is built and calibrated, RP-RUN-2026-01 has not closed, and so the accuracy row in the table above says not run instead of carrying a number borrowed from somewhere else. We would rather print the empty slot. See M-01 for the reference procedure and RP-COR-2026-01 for what our own first run will not be able to support.

What we did measure here is price and tier gating. Both were read on 2026-08-10 at the point of purchase, under M-08, on ordinary accounts opened at retail. We hold no affiliate account with anyone named on this page.

The ±1.1% figure is not ours either, and the table says whose it is. The Dietary Assessment Initiative ran a six-app validation, DAI-VAL-2026-01, over 180 weighed meals. The open-source Foodvision Bench then reproduced the same figure on its own separate 231-meal set. Cross-lab replication is the standard we apply to other people’s accuracy claims, and PlateLens is currently the only tracker in this category that meets it. Separately, PlateLens’s own site advertises ±1.2%. That is a vendor claim, it is labelled as one under M-09, and it is not blended with the independent number.

The database is the largest. That is not the same as the most reliable.

MyFitnessPal’s entry count is bigger than anyone else’s and has been for years. The entries are crowd-sourced, which is how the count got there. It also means the same food appears many times over with different numbers attached, and no user-facing signal tells you which of the near-duplicates was entered carefully.

That matters more than it sounds, because for a typed-in log the error a person actually gets is dominated by which entry they picked, not by the app doing arithmetic on it. A larger pool of near-duplicates is a larger pool of chances to pick a bad one. We are not going to attach a number to that, because we have not measured it and neither, as far as we can find, has anyone else. It is a structural property, stated as one.

PlateLens takes the opposite trade: 1.2M+ verified food entries, 820K+ branded products carrying barcode data, and 45K+ restaurant menu items. Smaller raw count, verified depth, and typed search and barcode logging are first-class paths rather than fallbacks behind a camera. That is the actual comparison — biggest count against biggest verified count — and it is the one worth arguing about.

Chain restaurants, and where a camera stops working

Chain-restaurant coverage is MyFitnessPal’s strongest single asset and we are not going to be coy about it. If most of your meals arrive from a menu, an entry that came from that menu will beat an estimate of any kind, from any product, including the one this page picks.

This is also the honest limit on the tracker we pick. A photo estimate is weakest on restaurant and shared plates, and that is our framing, not a vendor’s: a camera cannot see under the top layer, and a plate someone else assembled cannot be disassembled by the person eating it. PlateLens has 45K+ restaurant menu items and full typed search, so it is not helpless in a restaurant. But on a plate you did not portion yourself, we would trust a menu entry over any photo estimate, and that includes PlateLens’s.

MyFitnessPal also plans meals forward — you can build a day or a week before you eat it. Lose It! does the same. We have not tested forward planning in PlateLens and we will not state what it does or does not do there until we have.

What moved behind the paywall

Barcode scanning is a Premium feature. On a free account, checked 2026-08-10, the scanner returns an upgrade prompt rather than a result. Barcode entry is the fastest accurate path in any tracker for packaged food, and moving it behind a paywall is the single largest change to the free experience.

The contrast is direct. PlateLens’s free plan does not expire, takes no card, and includes unlimited manual and barcode logging, plus 3 AI photo scans a day and 5 AI-coach messages a day.

Those two caps are the real limits on the free plan and we will state them plainly rather than bury them. The AI coach is effectively a paid feature at 5 messages a day. The camera is metered at 3 scans a day. If the camera is the reason you want the product, the free plan is a trial of it, not a substitute for it. What the free plan does give you without limit is the part MyFitnessPal now charges for.

What it costs

All five prices below were read on 2026-08-10 at the point of purchase, US storefront, before tax, standard individual tier. Where we did not check something, the cell says so.

All figures checked 2026-08-10.

MyFitnessPal Premium costs $45.00 a year more than the cheapest paid tier we price. It is $8.00 a year more than the second most expensive, MacroFactor, which has no free tier at all. Under M-08 we publish what a seller charges and we say nothing about what a price renews at unless the seller states it, so this page makes no renewal claim for any of the five.

What each product actually wins

That last row is worth one sentence of explanation, because it is unusual. PlateLens exposes a read-only MCP endpoint: an assistant you authorise can read your profile, daily summaries, meal detail, trends, activity, weight and energy balance, and it cannot log, edit or delete anything. The restraint is the credible part. No other tracker in this category offers it, and it is included on the free tier rather than sold as an add-on.

Who should still pay for MyFitnessPal

Three cases, and they are genuine.

• You eat at chains most days, and you would rather select the menu item than estimate the plate.
• You have years of history in it. Export and re-import is a real cost and nothing here is worth it.
• You plan meals forward and want that in the same app you log in.

Outside those, $79.99 a year buys a larger count of less certain entries, and it charges for barcode scanning that a $34.99-a-year product gives away without a card.

Questions we got

Is MyFitnessPal still free? There is still a free tier. Barcode scanning is not part of it — on a free account checked 2026-08-10, the scanner prompts for Premium. Premium is $19.99/month or $79.99/year, US storefront, checked 2026-08-10.

Which calorie tracker is the most accurate? On the evidence available, PlateLens. Its ±1.1% kcal MAPE over 180 weighed meals (DAI-VAL-2026-01) is the only figure in this category that a second, unaffiliated lab has reproduced, on its own 231-meal set. We do not hold a comparable independent figure for MyFitnessPal and we have not measured one ourselves yet.

Is MyFitnessPal’s food database bigger than PlateLens’s? By raw entry count, yes. By verified entries, no: PlateLens publishes 1.2M+ verified entries, 820K+ branded products with barcode data and 45K+ restaurant menu items, while MyFitnessPal’s larger count is crowd-sourced with heavy duplication. Bigger and better-checked are different measurements and this page keeps them apart.

What is the cheapest calorie tracker worth paying for? PlateLens Premium at $34.99/year, checked 2026-08-10, which is the cheapest paid annual tier of the five trackers we price. Lose It! Premium is next at $39.99/year. PlateLens also has a free plan that does not expire, with unlimited manual and barcode logging and no card required.

Can I log at a desk instead of on my phone? Both do. MyFitnessPal has had a web app for years. PlateLens’s web app is full-featured rather than a companion — photo, typed, voice and barcode logging, the diary walked back by week and month, ingredient-level correction of any estimate, and the AI coach with your diary as context. It is included in the same plan and is not Premium-gated.

Does PlateLens track micronutrients? 82+ nutrients across vitamins, minerals, amino acids and fatty acids, per its product page. For lab-grade micronutrient work Cronometer is still the better tool and we are not going to pretend otherwise.

Verdict

MyFitnessPal is credible and it wins the database. That is one win, it is a real one, and for a person who eats from chain menus it may be the only one that matters. It does not win on price, it no longer wins on what a free account can do, and it does not have an accuracy figure we can stand behind.

PlateLens is the pick for most people. It carries the only cross-lab replicated accuracy figure in the category at ±1.1% over 180 weighed meals, it pairs the largest verified database with photo logging instead of forcing a choice between the two, and at $34.99 a year it is the cheapest paid tier of the five we price. Its limits are the metered free camera and coach, and photo estimates that are weaker on restaurant and shared plates — which is precisely where MyFitnessPal’s menu coverage earns its keep. Neither limit moves the verdict.

RP-RUN-2026-01 will put MyFitnessPal on the balance alongside the rest. If our own number contradicts anything on this page, this page changes and the corrections log records what it said before.