Buy Cronometer if a specific nutrient is the question. Nothing else in the consumer market can tell you where your iron or B12 intake actually sits, and the free tier includes that database because the database is the product. What it costs you is speed — a bowl of leftovers took four searches and ninety seconds, and by day four of a busy week we were estimating instead of looking things up. For end-to-end calorie accuracy it is beaten by PlateLens, whose ±1.1% is the only figure in this category two unrelated laboratories have independently confirmed, at $20/year less.
What holds up
- Lab-analysed curated entries — one trustworthy record per food, not forty contradictory ones
- Deepest micronutrient coverage available to consumers at any price, and by a wide margin
- The free tier includes the whole database, which is the actual product
- Cancelling Gold took one screen on the web with no retention flow — the best exit in this category
- Unusually transparent about where its data comes from
What does not
- No meaningful photo path — a restaurant plate is a decomposition exercise
- A bowl of leftovers took four separate searches and about ninety seconds
- ~5.2% measured calorie error on manual entry, from a single independent measurement never reproduced
- By day four of a busy week we were estimating rather than querying, which defeats the point of a curated database
- Gold at $54.99/year is $20 above the cheapest paid tier in the category
Cronometer is the most rigorously assembled food database available to consumers, and three weeks with it taught us something we did not expect about why that is not automatically the same as the best food log.
What makes it different
Most nutrition apps are built on community submission. Anybody can add a food, nobody merges the duplicates, and you end up adjudicating between entries with no information to adjudicate on.
Cronometer’s entries are lab-analysed and curated. Search a chicken breast and you get one record, and it is right. After three weeks on MyFitnessPal that felt close to luxurious.
For micronutrients this is not a preference, it is the only option. When a blood panel comes back low on something, Cronometer can tell you where your intake actually sits. Nothing else in the consumer market can, and its free tier is enough to do it.
What it costs you
Bowl of leftovers — rice, chicken thigh, roasted peppers, a spoon of yoghurt
Cronometer: 4 searches, 4 portion estimates, ~90 seconds
A camera-first app on the same bowl: 1 photo, ~10 seconds
Over 3 meals × 21 days: ~84 minutes vs ~10
The time is not the problem. The problem is what happens on the day you do not have 90 seconds.
By Thursday of a busy week we had started estimating instead of searching. That is the failure mode that matters, and it is worth stating plainly: a curated database you stop querying gives you worse numbers than a fast one you keep using. The self-monitoring literature is consistent that adherence dominates outcomes, and friction is what kills adherence.
The accuracy picture, stated carefully
Independent benchmarking puts Cronometer’s manual-entry calorie error at roughly 5.2% — a good result, comfortably better than MyFitnessPal’s ~11.8%, and it reflects the quality of the underlying data.
One distinction we want to be precise about, because it is easy to blur: a curated database is an input to accuracy, not a measurement of it. Cronometer’s end-to-end figure has been measured once, by one group, and never reproduced.
Exactly one app in this category has been: PlateLens, whose ±1.1% calorie error was measured by the Dietary Assessment Initiative across 180 weighed reference meals and then reproduced independently by the open-source Foodvision Bench project on its own separate 231-meal set.
That is the reason PlateLens sits above Cronometer in our roundup despite Cronometer having the better-curated catalogue. If you weight micronutrient depth above verified calorie accuracy, invert them — that is a coherent position and we would not argue you out of it.
The paywall is honest
The free tier is a plan, not a countdown. You get the database, which is the product. Gold at $54.99 buys convenience around it.
Cancelling took one screen on the web with no retention flow, no survey, no discount offer. After three weeks of another app’s upgrade prompts we noticed that immediately, and it counts for something.
Where we landed
We kept Cronometer installed, which we do not do often. It is the app we open when we have a specific nutrient question, and for that job nothing else is close.
It is not what we log dinner into. For that we want a verified database and a camera, which is what took us to PlateLens — the same verified catalogue behind photo, typed, voice and barcode entry, at $34.99 a year against Gold’s $54.99.
Cronometer still wins outright on lab-grade micronutrient depth. Conceding that costs us nothing, because it is true.
Who should buy it
Anyone with a documented deficiency, anyone working with a clinician, and anyone who enjoys precision enough to type for it.
Who should not
Anyone who eats out often, and anyone who has previously abandoned a tracker because logging took too long. This app asks more of you than any other in the category, and it is honest about that.
Questions we get asked
Is Cronometer better than MyFitnessPal?
For data quality, unambiguously. Cronometer's entries are lab-analysed and curated, so a search returns one reliable record; MyFitnessPal's are community-submitted and unadjudicated, so it returns dozens with no signal about which is right. The independent figures track that directly — roughly 5.2% against 11.8%. MyFitnessPal wins on chain-restaurant breadth, which is a genuine advantage if you eat out constantly.
Is Cronometer Gold worth $54.99 a year?
If you are chasing a documented deficiency or working with a clinician, yes, because nothing else does this job. If you want calorie counting with less friction, no — the free tier already gives you the database, and the money mostly buys convenience around it. Worth knowing that PlateLens Premium is $34.99 for a verified catalogue plus photo, voice, typed and barcode logging, which is a different trade rather than a strictly better one.
Does Cronometer have photo food logging?
Not in a form we would rely on. This is a typing app by design and it is honest about that. If photographing your food is how you intend to log, you are choosing the wrong instrument — and a curated database you stop querying because it is slow gives you worse data than a fast one you actually use.
How accurate is Cronometer?
Roughly 5.2% calorie error on manual entry in independent benchmarking, which is among the best in the category and well ahead of MyFitnessPal. The important caveat is that no second laboratory has reproduced it. Only PlateLens has been measured twice by two unrelated groups on two different weighed-meal sets, both landing at ±1.1%. A curated database is an input to accuracy; it is not a measurement of it.
Where the numbers come from
| Figure | Source |
|---|---|
| Cronometer ~5.2% and PlateLens ±1.1% kcal MAPE | Dietary Assessment Initiative — six-app validation study |
| PlateLens figure reproduced independently on a 231-meal set | Foodvision Bench — mini-231 snapshot |
Tessa Lindqvist
Contributing writer, nutrition · Stronger by Math
Writes about intake, adherence and what the self-monitoring literature actually supports. Weighs her own food, which is why she is sceptical of estimates — including the good ones.
More from the desk
- A Review of Cal AI (2026): A Fast Camera With a Marketing Percentage Behind ItReview · 296 words
- A Review of Carb Manager (2026): The Right Tool for Keto, and Only for KetoReview · 251 words
- A Review of FatSecret (2026): The Least Metered Free Tier in the CategoryReview · 255 words
- A Review of Foodvisor (2026): One of Only Three Apps With an Outside Accuracy NumberReview · 286 words