Everything that saves you typing is on the paid side of that line. I read MyFitnessPal's own help centre and its US App Store listing on 28 October 2026. The date each of those pages was last edited is in the table below.

I build a competing app, so the free tier gets credit here where it earns it.

Is MyFitnessPal still free in 2026?

Yes, and you can run it indefinitely without paying, as long as you find your food by searching. The company answers this on its own page: "The standard features of MyFitnessPal are free to use" (What is included in the free version, last updated 6 September 2026).

The download itself is free on the US App Store, where the listing shows ten in-app purchases (App Store). Ads pay for the rest, and MyFitnessPal says so plainly: "The core of MyFitnessPal is free, and it stays that way because ads help pay for it" (ad-free experience, 15 September 2026).

What changed since 2022 is not the price of free. It is how much now sits above it.

What do you actually get for free?

Eight rows, by MyFitnessPal's own count: food and exercise logging, weight and measurements, custom foods and recipes, viewing macros, percentage goals, diary sharing, partner app links, and the printable report.

The partner apps row is the one people miss. That is the row covering Apple Health, so nutrition still flows from a free account into Health, with caffeine, vitamin A and timestamps excluded (Apple Health connection and syncing, 24 September 2026).

The food database is free as well, and it is still the biggest in the category. Search a food by name, set the serving, done. Free MyFitnessPal runs that whole loop with no upsell in the way.

What is behind the Premium paywall?

Every shortcut that removes typing. Barcode Scan, Meal Scan, Voice Logging and the Nutrition Coach are all paid, and so is turning off the ads.

One oddity is worth flagging, because it is the kind of thing that makes people argue in forum threads. Voice Logging sits on the Premium features page under faster ways to log food, but it is not a row in the plan comparison table I counted, even though that table was refreshed two days ago (Premium features; plan comparison).

So the honest statement is that voice is Premium on the page that explains voice, and absent from the page that compares plans.

What does MyFitnessPal Premium cost in 2026?

$19.99 a month or $79.99 a year, as the headline purchases on the US App Store listing. The same listing shows ten Premium purchases at four different prices.

Two of those ten are yearly subscriptions at different prices. "MyFitnessPal Yearly Premium" appears twice at $79.99 and once at $49.99, alongside "MyFitnessPal Annual Premium" at $79.99 (App Store). Which one you are offered is not something the listing explains.

Premium+ I could not price from a primary source today. Its help page describes meal planning and grocery lists but gives no figure, and MyFitnessPal's pricing blog post answered every request I made on 28 October 2026 with a Cloudflare challenge. What I can state: Premium+ adds the Meal Planner, and it is sold in six countries, the United States, the United Kingdom, Canada, Ireland, New Zealand and Australia (Premium+, 20 September 2026).

What is the newest thing you need Premium for?

Nutrition Coach, the AI assistant MyFitnessPal added in 2026, which has been Premium from the start. Its help page was edited again on the day I checked, and it says Coach is "available to Premium and Premium+ MyFitnessPal members on iOS", in English, in six countries (Nutrition Coach, 28 October 2026).

Not everything new lands on the paid side. GLP-1 dose tracking, added for the United States and Canada, names no subscription anywhere on its help page and appears in no tier list (GLP-1 Support, 24 October 2026).

The pattern across four years is narrower than the arguments about it. Logging your food stays free. Logging it fast does not.

Is the free tier enough for you?

If you are happy to search the database by name, yes, and you should probably stop reading here. If what stops you logging is the typing, the free version is the one without the fix.

The US listing carries a 4.7 average from 2,384,277 ratings (App Store). That covers free and paid users together, so read it as popularity rather than a verdict on the paywall. The database and the Apple Health write, both free, are genuinely hard to beat.

Cronometer does barcode better, and gives it away: its own Gold page lists Barcode Scanner under Basic Functionality, the free column (Cronometer Gold, read 28 October 2026).

If the paywall is what pushed you out, I compared seven alternatives on price and free tier in MyFitnessPal alternatives, checked this month, and went through the in-app purchase list of each one in what is actually free without a subscription.

I build MealSaid, which is free with no ads and no in-app purchases, and you log by saying what you ate out loud. It is a much smaller thing than MyFitnessPal: no barcode scanner, no photo logging, no database to browse, iPhone only, iOS 26 and English only. For looking up a packaged food by name, free MyFitnessPal beats it outright.

How I checked this

Checked 28 October 2026. I read each MyFitnessPal help article through the help centre's public API, which also returns the last-updated timestamp printed in the table above, and counted the 33 rows of the Free, Premium and Premium+ comparison by hand (8 free, 21 Premium, 4 Premium+). The ten in-app purchases and the rating count came from the US App Store page for MyFitnessPal, read the same day.

Two things I could not verify. MyFitnessPal's pricing blog post is behind a Cloudflare challenge, so the Premium+ price is absent from this post rather than copied from a third party. Help centre pages return 403 to scripted fetches in a browser-less setting, so the links above open fine for you and may look broken to a link checker.

Corrections to ohmbrock42@gmail.com.

Apple Health accepts 39 nutrition fields (Apple developer documentation). I read each company's own documentation on 28 October 2026 and counted how many of those 39 the text actually names: MacroFactor 0, Cronometer 0, Lose It 1.

I build a calorie tracker, so I had to go and find this out for my own app. Here is each one, with the page it comes from and the date that page was last edited.

Does MacroFactor write nutrition to Apple Health?

Yes, and reads it back. MacroFactor's integrations page carries a five-row table, and the Apple Health column says Yes for Read Nutrition, Write Nutrition, Read Weight, Write Weight and Read Steps.

The catch is that the table is a PNG. Nothing in the page text says MacroFactor writes nutrition anywhere, so a search engine or a chatbot reading that page comes away with no answer (MacroFactor help, Integrations, read 28 October 2026).

The oldest primary statement I found is a March 2022 product update on MacroFactor's own site: "You can now export your nutrition data to Apple Health and Google Fit" (MacroFactor, March 2022).

Reading came later. The version 2.9.3 notes, dated 26 August 2024, added import of the day's calories, macros and micronutrients from Apple Health (release notes). MacroFactor ranks sources, with manual entries first, then its own food log, then Apple Health, so an imported total will not overwrite a day you logged yourself (Integrations).

That ranking rule is the best-documented conflict policy of any tracker I have read, and MacroFactor deserves the credit for it.

Does Lose It write nutrition to Apple Health?

Yes, and its help page is the only one of the three that lists what it sends in plain text. Weight, Food and Exercise go to Apple Health for everyone.

Premium users who have set a corresponding custom goal also send nutrient goals, blood pressure, blood glucose, BMI and body fat (Lose It help, Using Apple Health With Lose It!, last updated 1 April 2026).

Coming back the other way, Lose It reads food only if you grant it Dietary Energy, and it marks those entries: "Foods written from Apple Health to Lose It! will have a heart icon" (same page). That is the single Apple Health field name any of the three documents use.

One limit worth knowing before you switch apps. Lose It cannot backdate the connection, so historical data in Health will not flow in.

Does Cronometer write nutrition to Apple Health?

Yes, but its Apple Health page no longer says what. The page was rewritten on 27 October 2026, and what it covers now is importing activity, heart rate, sleep and backfilled days into Cronometer.

On writing, the live version says only that you can "manually select what data Cronometer can read and write into Apple Health" (Cronometer support, Apple Health and Apple Watch, last updated 27 October 2026).

I quoted an earlier version of that same page on 20 October 2026 in our Apple Health comparison, where it gave a count of 13 nutrients plus cholesterol. That sentence is gone from the version live today, and I cannot verify the number any more, so I am not repeating it.

Cronometer's Android page is a different story. Its Health Connect article lists 12 biometrics plus Food and Water as exports, item by item (Cronometer support, Health Connect, last updated 30 September 2026).

What does nutrition mean in Apple Health?

It means up to 39 separate fields, and an app picks its own subset. Apple groups them as 10 macronutrient fields, 13 vitamins, 8 minerals, 6 ultratrace minerals, water and caffeine.

Every one of those is a separate permission. An app that asks for dietaryEnergyConsumed and nothing else is honestly described as writing nutrition to Apple Health, and will still leave your protein history empty (Apple developer documentation, read 28 October 2026).

Which fields does each app actually name?

Between them, one. Lose It names Dietary Energy, and only in the context of reading food in.

Two of those rows are worse than they look. MacroFactor's answer exists but is unreadable to anything that cannot see a picture, and Cronometer's answer existed eight days ago and does not now.

Why does the missing field list matter?

Because the next app you try can only import what the last one wrote. Apple Health is the handover, and the handover is field by field.

Say you spend a year in one tracker, then move. If that tracker only ever wrote dietary energy, your new app sees calories and nothing else, however carefully you logged your macros. There is no way to find this out from a marketing page.

It matters for double counting too. Two apps writing the same meal give you a doubled day in Health, which is the single most common Apple Health complaint I see, and the fix is picking one source per field.

MealSaid, my app, writes five fields: dietary energy, protein, carbohydrates, total fat and fiber, plus workouts. It writes nothing else, so no vitamins and no sodium, and it does not read food from Health at all. It is iPhone only, needs iOS 26, is English only, and has no barcode scanner. If you want vitamins and minerals in Health, use Cronometer: its own page lists 84 nutrients in the free Basic column (Cronometer Gold, read 28 October 2026).

How I checked this

Checked 28 October 2026. I counted Apple's nutrition type identifiers from Apple's own documentation page and got 39 dietary fields across six groups. I read the MacroFactor, Lose It and Cronometer help pages the same day, and for the two Zendesk help centres I pulled each article through the public help centre API so I could record the exact last-updated timestamp rather than trusting a byline.

MacroFactor's integrations table is an image, which I downloaded and read directly; it has five rows and the Apple Health column reads Yes in all five. MacroFactor's help pages carry no last-updated date at all.

What I could not verify: the number of nutrients Cronometer exports to Apple Health, which its page no longer states, and whether MacroFactor writes individual micronutrient fields or only calories and macros. Both companies could settle this with one list.

Corrections to ohmbrock42@gmail.com.

The low bar matters more than it sounds. In an 8 week study of 90 adults using a food app, every week with three or more days logged was worth 0.23% of body weight, while recording at least half of your calorie goal predicted nothing at all (Payne et al., 2022).

I read six papers for this on 27 October 2026 and pulled six rules out of them. Where a number comes from a review rather than the original paper, I say so.

What are the six rules?

Hang the log on something you already do early in the day, and keep the bar at three days a week. Each rule below comes from one study, with its sample size.

How long before logging feels automatic?

About two months for a typical person, and anywhere from a few days to most of a year for a real one. Most of the useful information is in that spread.

Singh and colleagues searched six databases and found 20 experimental studies covering 2,601 people (Healthcare 2024). Only four of those reported how long a habit took to form.

Keller 2021 put healthy eating habits at a median of 59 days with a range of 4 to 335, and only 23% of participants got there at all. Lally 2010 reported a median of 66 days, Lally 2008 a mean of 91 days, and Fournier 2017 a mean of 106 days for morning stretching against 154 for evening. All four figures come from the 2024 review, which I read in full; I could not open the Lally papers themselves.

The 21 day figure that every habit article repeats is in none of them.

Does it matter when in the day I log?

Earlier is better, and immediately after something you already do is better still. Two studies point the same way from different behaviours.

Fournier's stretchers took 48 days longer to reach automaticity in the evening than in the morning, as reported in Singh's review. Mornings have fewer competing events, so the cue fires before the day can bury it.

Judah and colleagues gave all 50 participants the same motivational session, then told half to floss before brushing and half after. Those who flossed after brushing had stronger habits and flossed more often at the 8 month follow up (Judah et al., 2013).

So pick the end of an existing routine rather than a time. "After I pour the first coffee" beats "in the morning".

Why do reminders stop working?

Because an alarm trains you to answer the alarm, not to notice the meal. That is the uncomfortable result in the one study that compared the two directly.

Stawarz, Cox and Blandford ran a 4 week cue study and reviewed 115 habit apps for CHI 2015. Reminders supported repetition but "hindered habit development", while "event-based cues led to increased automaticity", and positive reinforcement did nothing (Bristol repository record).

One 4 week study is not a law, and a reminder you act on beats a habit you never form. Our reminder comparison found that at best 19.05% of logging reminders were tapped, which is about what you would expect from a cue with no event attached.

The practical version of an event cue is an if then sentence written in advance. Across 23 eating studies, those plans had a Cohen's d of 0.51 for adding healthy foods and 0.29 for cutting unhealthy ones (Adriaanse et al., Appetite 2011).

Is it better to log everything, or to log something every day?

Something, every few days. Logging roughly on three days a week predicted weight loss; logging your full calorie goal did not.

Payne and colleagues followed 90 employees with overweight or obesity for 8 weeks on Calorie Counter by FatSecret (Obes Sci Pract 2022). They scored three kinds of adherence: frequency, consistency (weeks with three or more days logged) and completeness (weeks recording at least half the calorie goal).

Consistency and frequency both predicted weight change, at 0.23% of body weight per consistent week and 0.02% per extra 1% of days. Completeness missed significance at p = 0.062 (regression results).

Participants averaged 4.4 consistent weeks out of 8 (Payne 2022). Running their slope out, the gap between 2 consistent weeks and all 8 is 1.38% of body weight, about 1.3 kg at this group's mean weight of 94.9 kg.

Two honest caveats, both from the same paper. Among the 50 people who finished the full 8 weeks, none of the three measures reached significance, and 89% of the sample backlogged at least one day rather than logging it live (results section).

How many times a day should I log?

Two or three, in small visits. Minutes spent did not separate the people who lost weight; the number of separate visits did.

Harvey and colleagues tracked 142 people through a 24 week online weight programme. Time spent self monitoring fell from 23.2 minutes a day in month 1 to 18.1 in month 3 and 14.6 in month 6, and 65.5% were still recording anything by then (Obesity 2019).

At month six, those who had lost 5% or more of their weight opened the journal 2.4 times a day against 1.6 for the rest, and the 10% group 2.7 against 1.7, both at p below 0.001 (Harvey et al., Obesity 2019). The authors' own reading is that successful participants logged "at least two to three times a day".

Average those three figures and you get 18.6 minutes a day, which over 24 weeks is about 52 hours of typing. That is the real price of the method, and it is why the effort side matters as much as the cue side.

Where does MealSaid fit, and where does it fail this test?

It cuts the cost of one entry; it does not give you a proper event cue. I built it, so take the second half seriously.

MealSaid calls your phone at times you choose and asks what you ate. That is a clock cue, the exact kind Stawarz found builds repetition rather than automaticity, and I am not going to pretend otherwise.

What it does help with is the three days a week bar. You say the meal in one sentence and it lands in Apple Health, or you say it to Siri with the app closed, which is how you attach a log to a real event like putting the plate in the sink. There is more on the mechanics in how MealSaid works.

It has no barcode scanner, no photo logging, no food database search and no way to import food other apps wrote to Apple Health. It is iPhone only and needs iOS 26. It understands English only. If you want a database you can audit, Cronometer is the better tool and this is not a close call.

For why the quitting happens in the first place, I went through seven dropout studies in why people quit calorie tracking.

How I checked this

Checked 27 October 2026. I read the Singh 2024 review, the Payne 2022 paper and the Harvey 2019 paper in full on PubMed Central, the Judah 2013 abstract on the authors' own repository at LSHTM, the Stawarz 2015 abstract on the University of Bristol research record, and the Adriaanse 2011 abstract through its publisher page.

Four numbers are second hand and labelled as such: the Keller 2021, Lally 2010, Lally 2008 and Fournier 2017 formation times, all taken from Singh's review because the original papers sit behind publisher walls that refuse automated access. The 52 hour figure and the 1.38% gap are my own arithmetic on the published means and regression slopes, not findings from the papers.

Nothing here is from MealSaid's own data. I have not published an adherence study of my own yet. Corrections to ohmbrock42@gmail.com.

I read every one of those App Store listings and help pages on 27 October 2026. The listings came from Apple's own lookup service, which also gives the date each one was last changed, and that date turned out to matter.

One app contradicts itself. Lose It's listing was updated on 26 October 2026, the day before I read it, to claim whole meal voice logging, while its own help article from August still describes voice as a way to search for one food.

Which iPhone apps can you actually talk to?

Eleven of 21, and MealSaid makes twelve. Nine of those twelve, mine included, say the app takes a whole spoken meal; ten apps say nothing about voice at all.

Prices are in our voice input comparison rather than repeated here, because they move faster than the features do.

What happens after you stop talking?

Two things, depending on the app: a list of matched foods you approve, or a search result you still pick from. Nothing in this field saves your words without showing you a guess first.

MyFitnessPal is the plainest about it. Its help article says "Voice Logging lets you say what you ate and get it into your diary", then walks you through tapping each food to fix the serving before tapping Log (MyFitnessPal, 21 September 2026).

Cronometer describes the same shape, as a feature that lets you "dictate or type out a description of the food you are eating" and returns matched foods to review (Cronometer, 26 October 2026).

Two apps claim voice and leave it there. Carb Manager's Premium page lists "Say (voice logging)" with no explanation, and BiteSnap's listing calls itself a voice enabled food tracker and stops.

MacroFactor is honest that the microphone is incidental. "Just type (or speak) a description of your food or meal", says its logging guide, which is the keyboard doing the speaking and the search doing the work (MacroFactor).

Which apps take the meal without opening the app?

Cronometer on Gold, MacrosFirst on Premium, and MealSaid. All three go through a Siri phrase, so the screen never has to come up.

Cronometer's Siri article, updated 24 October 2026, walks you through turning on its shortcut and then saying "Log food in Cronometer", after which Siri asks what you ate (Cronometer).

MacrosFirst does the same for Premium users, "all through Siri without ever opening the app", and handles serving size inside the Siri conversation (MacrosFirst).

Everyone else puts the microphone behind a screen. I counted the actual touches for each route in the hands free post.

Why do two of Lose It's own pages disagree?

Because the App Store listing was rewritten on 26 October 2026 and the help article has not caught up since 25 August. That is 62 days between two pages from the same company saying different things.

The listing now puts "Say 'I had 2 eggs, toast with butter and jam' to log your meal" under Premium plan features (Lose It, US App Store).

The help article still tells you to "use Say It to search for a food by voice", then pick the food, the serving unit and the quantity (Lose It help, 25 August 2026).

I am taking the listing as the newer fact and the help article as the one to watch. If you are paying for Lose It Premium to say whole meals, try it before the renewal lands.

Which apps do not let you talk at all?

Cal AI, Yazio, Noom, FatSecret, Nutritionix Track, Simple, Fastic, Zero, AteMate and MyPlate. None of them mentions voice or Siri on any page I could open.

Cal AI is the notable one, because it sells itself on effortless logging. Its listing of 20 October 2026 offers a photo of your meal and a typed description, and never a spoken one (Cal AI, US App Store).

FatSecret belongs in this column with an asterisk. Its listing of 6 October 2026 says nothing about voice, and its help centre returns 403 to every automated fetch I tried, so I am not going to describe a page I could not read.

BiteSnap sits at the other edge. Its listing calls it a "voice-enabled food tracker" and stops there, and that listing has not changed since 10 October 2025 (BiteSnap, US App Store).

Which one should I talk to?

If you already pay for a tracker, use the voice feature you bought; if you are starting fresh, decide first whether you want Siri or an in app microphone. That single choice removes most of the list.

For Siri with the app shut, the paid answer is Cronometer Gold, and it has the deepest nutrient data of anything here. Cronometer is the better tool if you want numbers you can audit, and I say that as a competitor.

MealSaid is my free answer to the same question. You say what you ate, it estimates calories, protein, carbs, fat and fiber and writes them to Apple Health, and a Siri phrase does it with the app closed.

It has no barcode scanner, no photo logging, no food database to browse and no restaurant menus. It is iPhone only and needs iOS 26, and it understands English only. Wrong app for a household that logs packaged food by barcode.

How I checked this

Checked 27 October 2026. I built the list by running US App Store searches for the names people use for calorie apps, keeping every general food or calorie tracker that came back, then reading each listing through Apple's lookup service and each company's own help centre.

Help centre dates come from the Zendesk timestamps on the MyFitnessPal, Lose It and Cronometer articles, read through those help centres' own article endpoints because the web pages return 403 to scripts, and from the published date on the Carb Manager page. MacroFactor and MacrosFirst show no date on their help pages, so those two rows say so rather than guessing.

Two gaps. FatSecret's help pages refused every request, so FatSecret sits in the no voice column rather than being described from a page I could not open. I did not install these apps or time them, so every row here is documentation reading, not a hands on test, and my own app is the only one I can speak about from use.

Corrections to ohmbrock42@gmail.com.

I counted the touches in each app's own step list on 26 October 2026. MyFitnessPal's Voice Log takes five touches from a locked phone to a saved entry, Cronometer's in-app voice logging takes five, and Lose It's Say It takes eight.

None of them work through the car screen, mine included. Apple's CarPlay category list has no slot for food logging.

What counts as hands-free?

Two moments decide it: whether you can start without touching the phone, and whether you can finish without touching it. Most voice features only solve the middle.

Turning speech into text is the easy part. Every app below does that well enough once you are already on its logging screen.

The cost sits at the two ends. You tap your way in, then you tap to confirm the match, and both taps want a dry finger and a look at the screen.

Which apps log a meal with zero touches?

Three, and all three use a Siri phrase instead of a button inside the app.

Cronometer is the one to look at if you want this on a mature tracker. Its article says Gold iOS users can log "without having to open the app", and that a banner confirms the item (Cronometer support, updated 24 October 2026).

MacrosFirst does the same for Premium users and handles serving size and meal in the Siri conversation, "all through Siri without ever opening the app" (MacrosFirst help). Its Log Food shortcut rides on AI Search, which its page calls a Premium feature.

Whether a phrase survives a locked phone is a choice the developer makes, not luck. Apple's App Intents framework has an authentication policy per action, and the alwaysAllowed case lets an action "run at any time, including when the device is locked" (Apple developer documentation).

Why does the tap after you speak matter?

Because that is the tap you cannot make with oil on your fingers. It is also the one that pulls your eyes back to the screen.

MyFitnessPal's iOS steps are open the app, tap the plus, tap Voice Log, speak, tap the blue arrow, tap Log. Voice Log is Premium, English only, and needs app version 24.35.0 or higher (MyFitnessPal help, updated 21 September 2026).

Cronometer's in-app route is shorter at five: the plus menu, Voice-Logging, speak, Match Foods, Add To Diary (Cronometer support, updated 24 October 2026).

Lose It is the longest of the group. Its article runs Log, Add Food, Say It, pick the food, pick a serving unit and quantity, Add or Save, then Done, and notes that Say It is English only (Lose It help, updated 25 August 2026).

MacroFactor's Describe accepts speech, in one parenthesis: "Just type (or speak) a description of your food or meal" (MacroFactor help). I found no Siri phrase for it in their help centre.

I searched the Lose It and MyFitnessPal help centres for "Siri" on 26 October 2026 and both returned zero articles. For a route comparison that includes widgets and the Lock Screen, see the post on logging without opening an app.

Can I log calories while driving?

Not through the car screen. Apple's CarPlay page names eleven kinds of app, and food logging is not one of them.

Apple's wording is about tasks you do on the way somewhere: EV charging, fueling, parking, public safety and quick food ordering "can help people accomplish important tasks without taking out their phone" (Apple CarPlay for developers). Ordering a burrito is a category. Recording that you ate one is not.

That leaves Siri on the phone, and it is worth being honest about what that costs. AAA had 120 drivers use the built-in systems of 30 model year 2017 cars, and rated 23 of the 30 at high or very high demand (AAA Foundation, 2017).

The useful detail for us is the direction. Voice lowered visual demand against the other modes, but the interactions ran longer, and the longer time cancelled the gain. A one-sentence log is near the short end of that range; anything that makes you read a list of matches is not, and should wait until you park.

What works when my hands are covered in oil?

A spoken Siri phrase, and after that the accessibility settings. Cooking is the easier case because you can look at the screen, you just cannot touch it.

Four things to set up once, in this order:

1. Settings, then Apple Intelligence & Siri, then Talk to Siri, so the phrase works without the side button. Cronometer's article walks through the same screens.
2. Settings, Accessibility, Siri. Cronometer's page points here for pause time and for the Always Listen option, which it says lets Siri listen while the phone is locked.
3. AirPods, if the phone is across the kitchen. The phrase is the same and the microphone is nearer your mouth.
4. For an app with no Siri phrase at all, Voice Control. Apple's page describes speaking commands "just like you would perform an action by touch", and the Show names command puts labels on everything tappable (Apple Support).

Voice Control is the slow option because you say the name of every button, and it is the only one that works on apps that shipped no voice support. It was built for people who cannot use a touchscreen, which is also the situation you are in with flour on both hands.

Where MealSaid is the wrong answer

It is a talking app with nothing behind it to browse. My own gaps, in the same spirit as the table.

• No CarPlay app, for the reason above, and no Android.
• No barcode scanner, no photo logging, no food database to search and no restaurant menus.
• iOS 26 and English only, and the phrase needs a connection, because the understanding happens on our server rather than on the phone.

Cronometer is the better choice if you want a verified database and a confirmation step before anything lands in your diary. Its own note recommends reviewing Siri-logged foods the next time you open the app, which is a fair warning that every app in this category should print.

How I checked this

Checked 26 October 2026. I read each company's own help article and counted the touches in its numbered steps, treating "open the app" as one touch and ignoring permission prompts and anything a saved meal would let you skip. The Cronometer, MyFitnessPal and Lose It articles are dated inside their help centres and I have quoted those dates above. The MacrosFirst and MacroFactor pages carry no visible date.

Two things I could not verify. Lose It's App Store description no longer mentions Say It at all, so I am not repeating the Premium label I used for it in an earlier post, and its logging article does not say which tier Say It belongs to. Neither Cronometer nor MacrosFirst states whether its Siri action runs on a locked phone, so those cells say "not stated" rather than "no".

I did not install the competitors for this post, and I did not time anything on a real phone; these are document counts. Corrections to ohmbrock42@gmail.com.

I counted those from each company's own help page on 26 October 2026. Then I read the research on why decisions, rather than taps, are the part that breaks.

One thing up front. The research below is about populations, not about you, and none of it shows that ADHD causes anything about weight.

Is there a calorie tracker made for ADHD?

No, and I would be wary of one that said it was. None of the apps I checked claims a clinical result, and none has been tested in an ADHD population.

What you can compare is friction, and four things carry most of it.

1. How many separate choices one entry costs.
2. Whether anything reaches you before the meal is forgotten.
3. Whether a missed day or a bad number punishes you when you open the app.
4. Whether you can log at 11pm for a lunch that happened at 1pm.

The numbers below cover the first two, which are the ones I could count from documents.

What does the research show about ADHD and eating?

A steady association with weight, a stronger one with eating disorders, and no evidence of cause in either direction.

The obesity figure comes from Cortese and colleagues, who pooled 42 studies covering 728,136 people. For adults the odds ratio was 1.55 (95% CI 1.32 to 1.81), and pooled obesity prevalence was 28.2% in adults with ADHD against 16.4% in those without (Am J Psychiatry 2016).

The same review reports that "individuals medicated for ADHD were not at higher risk of obesity", which is a useful brake on any story that treats this as a willpower gap (Am J Psychiatry 2016).

The eating disorder numbers are larger and matter more for how you use a tracker. Nazar and colleagues pooled 12 studies (4,013 people with ADHD against 29,404 controls) and found odds of 3.82 for any eating disorder, 4.13 for binge eating disorder, and 2.57 the other way round in 5 studies of eating disorder cohorts (Int J Eat Disord 2016).

That is my reason for saying something uncomfortable about my own category. If counting makes you restrict, a calorie app is the wrong tool, and the odds above say that applies to more people here than average. For scale, the CDC estimated 15.5 million US adults (6.0%) had a current ADHD diagnosis in 2023, from 7,046 interviews (MMWR, 10 October 2024).

How many decisions does one food entry take?

Six in Lose It, five in MyFitnessPal, four in Cronometer, one if you say the meal.

I counted a decision as any point where the app's own instructions ask you to pick between alternatives for this one food. Tapping a fixed button is not a decision. Picking which search result is your yoghurt is.

Lose It's article runs meal, tab (Search, My Foods, Meals or Recipes), search words, food, serving unit, quantity (Lose It help, updated 25 August 2026).

MyFitnessPal's runs meal, search words, result, serving size, number of servings (MyFitnessPal help, updated 23 October 2026).

Cronometer is the lightest of the three at four, because meal and quantity live on one screen: search words, food, serving size, diary group (Cronometer support, updated 25 October 2026). Cronometer does this better than the two bigger apps, and it is free for that route.

This is a different count from the one in my post on Cronometer's step list, which counted steps. Steps include the taps you do on autopilot. Decisions are the ones that need your attention back on the meal.

Why do decisions matter more than taps?

Because every decision asks you to hold the meal in mind while you answer something else. Lose the thread between "chicken burrito" and "which of these burritos", and the entry does not happen.

Cordeiro and colleagues surveyed 141 food journalers (47 active, 94 lapsed) alongside 5,526 forum posts. Their line on missed entries is blunt: "Simply forgetting to journal is a major barrier", and it was the most cited reason respondents had missed at least one. Asked what made journaling hard, 31 called it too much effort, 27 called it time consuming and 16 called it tedious (CHI 2015).

The same paper describes the other failure: people who had a bad day stopped opening the app, because red numbers read like a verdict. No app I checked lets you turn that off.

Patel's trial is the one that moves me most. Eighteen people logged every food in the Fitbit app and twenty ticked a daily checklist of high calorie foods; median days logged were 49% against 97%, and weight change was the same, 3.4 kg against 3.3 kg (JMIR Form Res 2022). Thirty-eight people is a pilot, not a verdict.

Do reminders fix the forgetting?

Partly, and less than you would hope. Chen and colleagues gave 30 people four weeks of both prompt types, and context triggered prompts were followed by a log within the hour 21.77% of the time against 17.32% for fixed-time ones (CHI 2024).

Read that as the ceiling for a banner. Four in five prompts went unanswered in the better condition, because a notification still asks you to go and do the six decisions. That study ran on Android phones, so treat the size of the gap as a direction rather than a number for iPhones.

MealSaid's answer is to make the interruption harder to dismiss: at times you pick, your phone rings like a normal call and a voice asks what you ate. Whether that beats a banner over a month is exactly the test I have not published yet, and I am not going to claim it until I have. The longer version of the dropout research is in why most people quit calorie tracking.

Where MealSaid is not the answer

It is one sentence per meal and nothing else, which is a feature until it is not.

• No barcode scanner and no database you can browse, so a packaged food you cannot name is a guess.
• iOS 26 and English only, and the phrase needs a connection because the understanding happens on our server.
• It estimates from your sentence. If you want gram-level control and a full micronutrient profile, Cronometer is the better app and I would say so to anyone.

And the thing that no app covers: if logging tips into counting that feels compulsive, stop and talk to a clinician rather than switching apps. The Nazar numbers above are the reason that sentence is in this post.

How I checked this

Checked 26 October 2026. The decision counts come from each company's current help article, read on that date, counting only the choices the instructions name for a single food, and ignoring the shortcuts (saved meals, favourites, recents) that cut every count. The research numbers come from the PubMed abstracts and open PDFs linked above, read the same day, and each one carries its sample size in the table or the sentence.

What I could not do. There is no study of calorie tracking apps in an ADHD population that I could find, so the friction argument here is an inference from general logging research, not a tested claim about ADHD. I also have no adherence data of my own yet; the 30 day comparison of reminders against a scheduled call is running on one person, which is me, and that is a weak design I will label as such when it goes up.

Corrections and better papers to ohmbrock42@gmail.com.

The weight in Cronometer sits inside those four steps: seven category tabs, a Sort By menu, a database filter and a Better Alternative swap, all described on the same help page.

Cronometer has its own fast route, a spoken sentence, and it costs $59.99 a year.

I read five help centres and four App Store listings on 25 October 2026 and counted the steps each company publishes. I build a calorie tracker, so treat the second to last section as the biased one.

How many steps does Cronometer take to log one food?

Four, by its own count: tap the plus icon, tap Add Food, type the name, tap the food to open serving size and diary group.

That is the whole numbered list on Cronometer's Mobile Add a Food page (Cronometer, updated 24 October 2026).

MyFitnessPal publishes six steps for iOS (MyFitnessPal, updated 23 October 2026). Lose It publishes nine (Lose It!, updated 25 August 2026).

So on the published numbers, Cronometer is already the short one. I went in expecting the opposite and had to redo the chart.

So why does Cronometer feel slow?

Because step three hides a screen full of decisions, and the page documents every one of them.

The food search carries seven category tabs: All, Favorites, Common, Supplements, Brands, Custom and Restaurants. Above them sits a Sort By menu and a database filter that can be pointed at one source such as NCCDB or a combination.

Some brand foods are then tagged with a Better Alternative, a curated swap with fuller data, and a Multi-Add toggle turns the list into tick boxes with a Review Items step (Cronometer).

Every one of those is there for a reason. Cronometer even keeps a Crono-Hacks page for people who want more of this, with recipes built out of daily supplements and custom biometrics for nutrients it does not track (Cronometer, updated 4 October 2026).

That is a nutrition instrument. If you want one, nothing below replaces it.

Which app asks you for the least?

MyNetDiary, once you use its Describe box: five published steps, on a tier that is free and ad-free.

Four notes on that table. A step in a help page is not a tap on a phone, and favourites, recents and saved meals shorten every one of these counts.

Free means with ads in two cases. Cronometer puts "No Advertisements" on its Gold list (Cronometer, updated 24 October 2026), and MyFitnessPal says plainly that ads pay for its free core (MyFitnessPal, updated 15 September 2026).

Apple also never says what period MyNetDiary's $59.99 Premium purchase buys, so that row is the one price here I cannot date to a billing cycle.

MacroFactor calls its logger the "fastest food logger on the market" and writes its instructions as prose rather than a numbered list, so its six is my count of the actions in order (MacroFactor).

Can you just talk to Cronometer instead?

Yes, and both voice routes sit behind Gold at $59.99 a year.

The help page is blunt about it: "Gold users can take advantage of our AI Voice-Logging feature". You open the orange plus menu, describe the food, tap Match Foods, then tap Add To Diary (Cronometer, updated 24 October 2026).

The Siri route is the one to read before you buy. Cronometer's own getting started list runs to eleven sub-steps across three stages: install Shortcuts, enable Siri on the Cronometer shortcut, then turn on voice activation in iOS Settings (Cronometer, updated 24 October 2026).

MyNetDiary's Describe costs nothing for the first three tries, then needs Premium, and it is five steps including the review (MyNetDiary, updated 21 October 2026). Our post on what each voice feature actually means has the rest.

What do you lose by leaving Cronometer?

The data quality, and most of your history.

Cronometer's free Basic account already covers up to 95 nutrients, over a million food entries and unlimited barcode scanning (Cronometer). MyNetDiary's free tier matches it on nutrient count, at 108.

What it does not match is the bookkeeping. Cronometer shows a Data Confidence figure on a nutrient target, because a branded food often carries only the fifteen or so values a US label requires (Cronometer, updated 16 October 2026).

Then there is the history. MyNetDiary's own migration page tells you to move Cronometer data through Apple Health, and lists weight, exercise and other health data (MyNetDiary, updated 3 August 2025). Your food diary is not on that list.

If depth is why you are on Cronometer, read our MacroFactor and Cronometer comparison before switching to anything.

Which one would I use?

MyNetDiary, if I wanted a database and a free account. Mine, if I wanted the meal logged and forgotten.

MealSaid is the narrow option here. You say the whole meal and it writes calories, protein, carbs, fat and fiber to Apple Health, and at times you pick your phone rings and a voice asks what you ate.

What it does not do is most of this post. No barcode scanner, no food database to search, no photo logging, no micronutrients beyond those five fields, iPhone only on iOS 26, English only.

Anyone who tracks vitamins, minerals or a medical diet should stay on Cronometer. The case for a simpler app is for people whose diary is empty by Thursday, which is the failure mode the research keeps finding.

How I checked this

Checked 25 October 2026. I read eleven pages for the step counts: Cronometer's Add a Food, Voice Logging, Siri Voice Logging and Crono-Hacks articles, MyFitnessPal's logging article, Lose It's logging and pricing articles, MyNetDiary's Describe and free tier pages, and MacroFactor's two logging articles.

Prices came off each US App Store listing on the same day, where Cronometer shows Gold Annual at $59.99 and Gold Monthly at $10.99. Apple lists at most ten purchases per app, so treat those as a floor rather than a full price list.

The step counts are counts of published instructions, not timed runs on a phone. My own timed test across six apps is still unpublished, so MealSaid's one step in the table is a description of the flow and not a measurement. The Cronometer, MyFitnessPal and Lose It help centres return 403 to scripts, so those articles were read through each help centre's public API and the updated dates are the ones the API reports. Corrections to ohmbrock42@gmail.com.

Cronometer is the runner up with 5 home screen widgets and a Watch app that adds food, but its Siri route is locked behind Gold at $59.99 a year.

I checked what each company documents on 25 October 2026, not what the App Store screenshots suggest. I build one of these apps, so the second to last section is the one to discount.

Which free tracker does the most on an iPhone?

MyNetDiary, by a wide margin on documented iPhone features.

Its free tier page lists the barcode scanner, voice input, 108 nutrients, Apple Health sync, the Apple Watch app and Siri, describing the core version as "free and ad-free" (MyNetDiary, updated 15 October 2026).

The Watch app is the part other apps do not match. MyNetDiary says it logs foods, meals, water and quick calories from the wrist and can show any tracked nutrient (MyNetDiary).

That iPhone page is MyNetDiary selling MyNetDiary, so read its counts as the company's claim. Everything on the other four rows below comes from a help article instead.

Each Siri and Watch claim in that table is linked in the section that covers it.

MacroFactor is missing on purpose. Its App Store listing says it "does not offer a free subscription tier" and describes a seven day trial, so nothing it does on an iPhone is free (App Store).

How many widgets does each app actually have?

Sixteen for MyNetDiary, five for Cronometer, three each for Lose It and MyFitnessPal.

Cronometer names Energy Consumed, Energy Remaining, Fasting, Water Tracking and Streaks. The water widget adds and removes glasses from the home screen, and the Energy Remaining widget carries an add food button (Cronometer, updated 22 October 2026).

Lose It names three iOS widgets: Goal Ring Calorie Countdown, Macros Breakdown and Log Foods (Lose It!, updated 6 July 2026).

MyFitnessPal's widget page names Today's Calories and Today's Macros, then adds an Interactive Water Widget for iOS 18 or later, which is the only one of its widgets that writes anything (MyFitnessPal, updated 27 August 2026).

Two of those counts are floors. MyFitnessPal's page says there are "a few options to choose from such as" the two it names, and I counted what was named.

Can you log a meal with Siri for free?

Only a meal you already saved. Of the five apps here, open ended Siri logging is paid everywhere except MealSaid.

MyNetDiary's Siri page describes built in shortcuts for calorie checks, the daily deficit, water logging and My Meals logging, added from Add to Siri buttons on the matching screens (MyNetDiary). Say the phrase and a saved meal goes in; a sandwich you invented at lunch does not.

Cronometer does take any food by voice, and it is a Gold feature at $59.99 a year. The setup list runs to eleven sub-steps across three stages before the first log (Cronometer, updated 24 October 2026).

I could find no Siri phrase that logs food in MyFitnessPal or Lose It. Both put voice behind a button inside the app, which is the whole point of the post on logging without opening the app.

Does anything reach the Lock Screen?

MyNetDiary, with 12 lock screen widgets by its own count.

Cronometer, MyFitnessPal and Lose It all document home screen widgets only, and searching all three help centres for lock screen articles returned nothing about widgets or controls.

MealSaid has a Lock Screen control, which is a different iOS thing: a button in the control picker that opens the app already listening, rather than a widget that displays numbers.

Is the free Apple Watch app worth anything?

It depends what you call logging. Two of the four will take a food, two will take a number.

Cronometer's Watch app shows the Energy Summary and has an Add Food widget that lists your most recent items (Cronometer, updated 19 October 2026).

MyFitnessPal's Watch app lets you add water or quick add calories, and feeds step data back to your goal (MyFitnessPal, updated 27 August 2026). Lose It does not take food on the wrist at all: its Watch support is a free calorie bonus from resting and active calories, and no food entry is described (Lose It!, updated 9 September 2026).

That read figure matters more than it looks. An app that reads nutrition back can pick up meals you logged elsewhere, which is the test in our Apple Health comparison.

Which would I put on my own iPhone?

MyNetDiary first, then mine for the meals I would otherwise skip.

MealSaid is one surface done one way. You say the whole meal, it writes calories, protein, carbs, fat and fiber to Apple Health, and at times you pick your phone rings and a voice asks what you ate.

Against the table above it is thin. No barcode scanner, no food database, no Apple Watch app I can point you to, no nutrients beyond those five, iPhone only on iOS 26, English only.

MyNetDiary beats it on everything a feature list can measure, free, and anyone comparing what each free tier keeps should start there. The narrower pitch is for people who have installed three of these and logged in none of them.

How I checked this

Checked 25 October 2026. The widget, Watch and Siri claims come from each company's own help article or iPhone page, read on that date through the public help centre API where the site returns 403 to scripts. MyNetDiary's widget and Watch numbers come from its marketing page rather than a help article, so they are the company's claim, not a count I made from a phone.

I did not install five apps and photograph their widget pickers. A count of named widgets is a floor, since a company can ship one and not document it, and MyFitnessPal's page openly says there are more than the two home screen widgets it names.

MealSaid's row describes my own app. Its Siri phrase, Lock Screen control and widget are described in the earlier post, and the step counts behind the other apps' logging flows are in the Cronometer alternative post. Corrections to ohmbrock42@gmail.com.

I read both help centres and both US App Store listings on 24 October 2026. Every price and feature below comes off those pages.

I build a calorie tracker myself, so treat the last section as the biased one and the rest as what the documents say.

Which one costs less in 2026?

Cronometer Gold at $59.99 a year, against $71.99 for MacroFactor.

Cronometer's US App Store listing shows two Gold purchases, Gold Monthly at $10.99 and Gold Annual at $59.99 (App Store).

MacroFactor's help centre lists $11.99 a month, $47.99 for half a year and $71.99 a year for a single app, and the same three prices appear on its App Store page (MacroFactor).

There is a fourth MacroFactor price. A bundle covering the separate MacroFactor Workouts app is $89.99 a year, and anyone who subscribed before 1 January 2026 keeps Workouts free until 12 January 2027 (MacroFactor).

Is either one actually free?

Only Cronometer, and its free account is not a stub.

A Basic account covers all 95 nutrients, the whole database of over a million entries, unlimited barcode scanning, custom meals and recipes, and device syncing (Cronometer). The limits are the advertising and a seven day window on your nutrition history, with two charts in place of the full library.

Gold buys the ad-free view, long term charts and reports, the Oracle food suggestions, fasting, custom biometrics and the recipe importer (Cronometer support, updated 17 October 2026).

MacroFactor does not do free. Its App Store listing is unusually blunt about it: the app "has a free trial, but does not offer a free subscription tier", and the same listing describes a seven day trial (App Store).

MacroFactor and Cronometer side by side

MacroFactor costs more and hands you every logging method on day one. Cronometer goes deeper on nutrients and lets you stay free forever.

How many nutrients does each one track?

Cronometer says up to 95. MacroFactor's help centre documents 56, and I counted them one page at a time.

MacroFactor publishes a Nutrients collection with one reference article per nutrient it tracks. On 24 October 2026 that collection held 56 articles: 9 on calories and the macro breakdown, 11 amino acids, 9 kinds of fat, 12 vitamins, 10 minerals, and 5 others covering cholesterol, caffeine, alcohol, water and choline (MacroFactor).

Cronometer's Track Food page promises "up to 95 nutrients and compounds" (Cronometer), and its free tier page repeats the 95. Its own home page says 84 in the paragraph about Cronometer Pro, so their marketing does not agree with itself. Either way there is no list I could count, so 95 is their figure and 56 is mine.

Both warn you that the number is a ceiling, not what you will actually see. MacroFactor says most branded products do not report deep micronutrient data, so anything beyond calories, macros and "some major micronutrients" will be undercounted unless you log common foods (MacroFactor).

Cronometer hides some nutrients by default for the same reason, and warns that tracked values "may be significantly lower than they actually are" (Cronometer support, updated 14 August 2026).

So the real question is how many foods in each database carry a full nutrient profile. MacroFactor says 26,500 of roughly 1,360,000 searchable items are micronutrient complete, which is 1.9% (MacroFactor). Cronometer's lab analysed core is the NCC database at over 17,000 entries with data on 70 nutrients, plus USDA SR28 at over 8,000 (Cronometer support, updated 21 October 2026). Its whole database runs past a million entries and it does not say how much of that is lab analysed (Cronometer).

Cronometer does this part better. It also sells a Pro tier aimed at dietitians and research teams, which is not a market MacroFactor is chasing (Cronometer support).

Which one gets your food into Apple Health?

Both write. Only MacroFactor reads nutrition back in.

MacroFactor's integrations page spells out a priority order: manual entry, then its own food log, then Apple Health. A MacroFactor food entry for a day blocks the Health import for that day (MacroFactor).

Cronometer's Apple Health article covers activity, exercise, biometrics and sleep coming in, and tells you to pick read and write permissions, but nothing in it describes importing food logged elsewhere (Cronometer support, updated 22 October 2026).

One thing changed while I was writing. When I read that article on 20 October 2026 for our Apple Health comparison, it named the nutrient fields Cronometer writes. The 22 October version does not, and no other article in their help centre lists them now. Treat the write list as undocumented until they put it back.

Which one should you pick?

Cronometer if you care about micronutrients or want to pay nothing. MacroFactor if you want the coaching algorithm and do not mind the $71.99.

MacroFactor's pitch is its expenditure algorithm, which reads your weight trend against your logged intake and updates your targets. It needs food logged "at least four times per week, but ideally every day" to keep running (MacroFactor).

Cronometer's pitch is depth, and it charges for the fast logging paths; our voice input comparison has the detail on that.

Mine is the third option and the narrow one. You say what you ate and MealSaid writes calories and macros into Apple Health. If you forget, it calls your phone and asks.

It also has no barcode scanner, no food database to search, no Android app, no web app, no vitamins or minerals at all, and it needs iOS 26. If you want 95 nutrients, Cronometer Basic gives you them for nothing. You can read how MealSaid works before deciding.

How I checked this

Checked 24 October 2026. Prices came from the In App Purchases panel on each US App Store listing and from MacroFactor's own subscription article; Apple shows at most ten purchases per listing, so those are floors.

The 56 nutrient count is my own count of the articles in MacroFactor's public Nutrients collection on that date, so it will drift as they add pages. The 95 is Cronometer's claim on two of its own feature pages and I could not verify it item by item. Their home page says 84 for Cronometer Pro, and I did not find any page that reconciles the two.

Cronometer's help centre returns 403 to scripted fetches, so those articles were read through the public help centre API and their "updated" dates are the ones the API reports. Corrections to ohmbrock42@gmail.com.

Cal AI itself is not camera-only. Its site says "Snap a photo, scan a barcode, or describe your meal", so the description route exists (Cal AI). Two of those three routes still need the camera, and all of them need a subscription.

I read every price and feature below on each company's own site or US App Store listing on 24 October 2026.

Does Cal AI work without the camera?

Yes, through the describe route, but the app is built around the photo and says so.

The App Store listing gives three steps to use Cal AI, and step two is "Snap a photo of your meal" (App Store).

The feature Cal AI sells hardest is the one that cannot work without a lens: "your phone's depth sensor calculates food volume" (Cal AI).

The same listing carries a line in capitals saying food scanning analysis requires a subscription, so the camera-free route is not a free route either.

What does Cal AI cost?

Nobody can tell you, including Cal AI. Apple lists one purchase named Unlimited at five different prices between $2.99 and $29.99.

Cal AI's website has no pricing page. What it has is a FAQ about cancelling, a FAQ about refunds, a FAQ about restoring a purchase and a button offering a three day free trial (Cal AI).

The In App Purchases panel on its US listing shows Unlimited at $2.99, $5.99, $9.99, $19.99 and $29.99, plus a $0.99 Streak Restore (App Store). Apple never says what period any of them buys.

The top price is ten times the bottom one for a product with one name. You find out which one you were shown after the trial starts.

Which apps log a meal without a camera?

All six alternatives do, and so does Cal AI. Four of them let you type into a search box for free, and only MealSaid turns a whole spoken sentence into a meal at no cost.

Two rows need a footnote. MyFitnessPal's Voice Logging is Premium only, on iOS and Android, English only, app version 24.35.0 or higher (MyFitnessPal, updated 21 September 2026). And Lose It's Say It searches for one food at a time rather than taking a sentence (Lose It!).

Which free tier actually does the most?

MyNetDiary, and it is not close.

Its own feature table marks the staff verified database, unlimited search, the barcode scanner, voice logging, exercise and water tracking, up to 108 nutrient fields and the charts as free, with no ads and no account required to start (MyNetDiary).

The catch is what "voice logging" means there. Free voice logging finds foods; the version that turns a spoken meal description into structured diary entries sits under Premium Plus, at $14.99 a month or $99.99 a year (MyNetDiary).

Cronometer's free tier is the other strong one: all 95 nutrients, the whole database and unlimited barcode scanning, with ads (Cronometer). Its voice route costs $59.99 a year.

If free is the whole requirement, start with MyNetDiary and read what is actually free in each app before you install anything.

Is describing a meal more accurate than photographing it?

Nobody knows, and anyone who tells you otherwise is guessing.

Photos have been measured once properly. Researchers at the NIH weighed 102 meals to 0.1 g, photographed them and ran them through four apps; Cal AI came in 345 calories low per meal, the worst of the four, and Appediet came in 252 low (Healio). Our full write up of that study has the confidence intervals.

There is no equivalent test for spoken or typed descriptions. Not for MyFitnessPal, not for Cronometer, not for mine.

What I can say is where the error comes from in both cases: portion size. A photograph does not show the oil absorbed into the food, and neither does the sentence "chicken and rice".

Which one would I use?

MyNetDiary if I wanted a food database and a barcode scanner for nothing. MealSaid if I just wanted the meal logged and gone.

Cronometer Gold sits between them: $59.99 a year buys the spoken meal plus the deepest nutrient data of the six, which is covered in our MacroFactor and Cronometer comparison.

Mine is the narrow one. You say what you ate and MealSaid writes calories, protein, carbs, fat and fiber into Apple Health. At times you pick, your phone rings and a voice asks what you ate.

It has no camera features at all, which is the point of this post, but that also means no barcode scanner, no nutrition label scanner, no food database to search and no photo logging. It is iPhone only and needs iOS 26. English only, too. On raw features MyNetDiary beats it, free, and I would say so to anyone who asked.

How I checked this

Checked 24 October 2026. Prices came from the In App Purchases panel on each US App Store listing and from each company's own pricing or feature page where one exists. Apple lists at most ten purchases per app, so those ranges are floors.

Cal AI is the one app here with no published price. I did not start its trial to find out which of the five Unlimited prices I would be shown, so that gap stays open.

The MyFitnessPal, Lose It and Cronometer help centres return 403 to scripted fetches, so those articles were read through each help centre's public API, and the "updated" dates quoted are the ones the API reports. Corrections to ohmbrock42@gmail.com.

I checked this because the comparison posts I found were quoting barcode scanning as a free Lose It feature, and Lose It's own help centre says otherwise. Every row below comes from one of the two companies' own pages or their US App Store listing, read on 23 October 2026.

What does each one cost in October 2026?

The same: $79.99 a year. The difference is how many other prices each one shows you.

Lose It's membership article gives one number for Premium, "$79.99 USD / Year (Auto-renews annually)", plus a Lifetime unlock at $299.99, or $229.99 for an active Premium member (Lose It! help centre). New users get a 7 day Premium trial.

MyFitnessPal's US App Store listing shows ten purchases at four distinct prices: $9.99, $19.99, $49.99 and $79.99, with $79.99 attached to three separate yearly SKUs (App Store).

Lose It's listing shows ten purchases at seven distinct prices for subscriptions, from $9.99 to $79.99, plus two more named Premium Lifetime at $49.99 and $59.99 (App Store). Those two do not match the $299.99 in the help centre, and I cannot reconcile them.

Lose It and MyFitnessPal, line by line

Eleven things people actually ask about, each cell read off a MyFitnessPal or Lose It page on 23 October 2026.

Can you still scan a barcode without paying?

Not in MyFitnessPal since 1 October 2022, and Lose It is in the middle of making the same change.

MyFitnessPal states it flatly: as of 1 October 2022, Barcode Scan requires a Premium subscription (MyFitnessPal help centre). That one is old news.

Lose It is the new part. Its article on unlocking the scanner says Scan It and the nutrition label scanner are becoming Premium features, and that "some free members will still have access as we transition, while others will not" (Lose It! help centre, read 23 October 2026).

So your Lose It free account may still scan today and stop next month, and the roundups that list barcode scanning as a free Lose It feature are describing a state that is being retired. I had it that way in the MyFitnessPal alternatives table too, and that row needs correcting.

How much of MyFitnessPal is actually free?

Eight features out of the 33 in its own comparison table, by my count.

MyFitnessPal publishes a Free, Premium and Premium+ grid. I counted the rows on 23 October 2026: 33 features listed, 8 ticked for Free, 21 starting at Premium, 4 exclusive to Premium+ (MyFitnessPal help centre).

The eight free ones are the boring essentials: food and exercise logging, weights and measurements, your own recipes, macronutrient views, goals by percentage, diary sharing, partner app links and a printable report.

Lose It publishes no equivalent grid, which is why this count exists for one app and not both. Its membership article describes Premium in a sentence rather than a table.

Both free tiers are paid for by promotion. MyFitnessPal is unusually direct about it: the core is free and "it stays that way because ads help pay for it" (MyFitnessPal help centre). Lose It runs in-app pop-ups instead, and the toggle to switch them off is a Premium control (Lose It! help centre).

Which one fits Apple Health better?

Lose It, and it is not close, because it reads food back.

MyFitnessPal's Apple Health article gives a four-row sync table. Food goes one way only, out to Apple Health, minus caffeine and vitamin A and minus timestamps. Exercise, sleep and weight come in (MyFitnessPal help centre).

Lose It moves weight, food and exercise in both directions. Food written to Apple Health by another app shows up in your Lose It log with a heart icon, once you grant Dietary Energy read permission (Lose It! help centre).

That is the one row here with a real consequence. If you ever log from a second app, a watch or a Shortcut, Lose It will show you the total and MyFitnessPal will not. The full read and write matrix across six apps is in our Apple Health comparison.

Does either one remind you to log?

MyFitnessPal does, per meal, only when you have not logged. I could find no reminder feature in Lose It's help centre.

MyFitnessPal's reminders screen lets you add a notification per meal, and the setting is literally worded "Remind me if I haven't logged" (MyFitnessPal help centre). It is free, not Premium.

Searching Lose It's help centre for reminders on 23 October 2026 returned subscription renewal notices and fasting guides, nothing about logging prompts. Absence from a help centre is weaker evidence than a documented feature, so treat that cell as unconfirmed rather than proven false.

So which should you pick?

Lose It if your food total in Apple Health has to be real. MyFitnessPal if you want a logging reminder that costs nothing.

At an identical $79.99 a year, the paid tiers are close enough that the free tiers decide it, and both free tiers are now a search box with promotions attached. That is the honest state of play, and it is the reason this blog exists.

My own app, MealSaid, is the footnote here: free with no in-app purchases, you say what you ate and it writes calories and macros to Apple Health. It has no barcode scanner, no photo logging, no food database to search and no Android build. It needs iOS 26 and it is English only. For what is genuinely free across the wider market, see the free calorie counter breakdown.

How I checked this

Checked 23 October 2026. Every feature row came from a MyFitnessPal or Lose It help centre article read through each site's public Help Center API, because both sit behind a bot wall that returns 403 to scripted page fetches. Article update dates ranged from 1 April 2026 (Lose It's barcode notice) to 20 October 2026 (MyFitnessPal's cancellation article).

Prices came from two places: each company's own help article, and the In-App Purchases panel on each US App Store listing, which shows at most ten purchases. The count of 8 free features out of 33 is mine, made by counting rows in MyFitnessPal's published grid.

What I could not verify: MyFitnessPal's own pricing page, which returns a Cloudflare challenge to everything I tried, so there is no trial length in the table. I also could not reconcile Lose It's $299.99 Lifetime price with the $49.99 and $59.99 Premium Lifetime purchases on its App Store listing. Corrections to ohmbrock42@gmail.com.

I went looking for this because people keep asking me whether a bigger food database would make MealSaid more accurate. The honest answer is no, and the research is not subtle about it. Checked 23 October 2026. Every figure below links to the paper it came from.

How many calories a day is the gap, really?

Between 330 and 579 calories a day, in the three studies that measured real energy expenditure rather than asking people a second time.

Doubly labeled water is the reference method. You drink isotope-labeled water, researchers track how fast the labels leave your body, and that gives total energy expenditure without anyone reporting anything. In weight-stable people it is the closest thing to a true intake number.

Boushey and colleagues dosed 45 adults, had them photograph every eating occasion for 7.5 days, and had trained analysts estimate the calories. Measured expenditure averaged 2,932 calories a day. Reported intake averaged 2,353, "a resulting difference of 579 kcal/day" (Nutrients 2017).

Serra and colleagues ran 30 women through seven days of doubly labeled water in Zurich while they logged with SNAQ, a food-recognition app. The gap was 330 calories a day, or 12.8% low. A one-day 24 hour recall in the same 30 women was 543 calories low, 22.9% (Frontiers in Nutrition 2023).

Foster and colleagues checked Intake24, an online recall tool, against doubly labeled water in 98 UK adults aged 40 to 65. People under-reported by 25% on a first recall and 25% again on the average of three (Journal of Nutritional Science 2019).

Is it the app's database, or is it you?

It is you, by roughly thirteen to one. The database gap is 26 calories a day and the reporting gap is 330.

Evenepoel and colleagues took 50 people's four-day food records and ran them through MyFitnessPal and through Nubel, the Belgian national food composition database. Energy came out at 1,984 calories a day in MyFitnessPal against 1,958 in Nubel, with a correlation of 0.96 (JMIR 2020).

That is a 26 calorie difference, my own subtraction on their Table 1. Put it next to the 330 calorie gap Serra measured and the database accounts for about 8% of the problem.

The same paper found the database is not uniformly fine. Sodium came out at 1,293 mg a day against the reference 2,638, cholesterol at 55 mg against 242, and fiber at 15 g against 19 (Table 1). If you track sodium, that matters a lot; if you track calories, it does not.

Why does a 12.8% average still ruin a week?

Because the average is not the error you get. Serra's 95% limits of agreement ran from 1,504 calories under to 845 calories over (Table 3).

A Bland-Altman limit of agreement says where 95% of individual measurements fall. For the photo app, a given woman's day could plausibly be 1,504 calories low or 845 calories high, and the midpoint of 330 describes nobody in particular.

Foster's numbers say the same thing in percentages. The point estimate was 25% under, and the limits ran from 73% under to 68% over (Foster 2019).

Boushey put the same thing in people rather than intervals. Of the 45 adults, 53% were classified as underreporters, and that group was off by a mean of 1,000 calories a day (Boushey 2017).

Are the numbers on the menu any better?

On average yes, per item no. Of 269 restaurant foods burned in a calorimeter, 50 held at least 100 calories more than the menu claimed (Urban 2011).

Urban and colleagues bought food from 42 randomly selected quick-serve and sit-down restaurants in three US states and measured the energy by bomb calorimetry (JAMA 2011). The overall gap was 10 calories a portion, nowhere near significant.

Then the detail: 108 items (40%) were at least 10 calories over the stated value, 141 (52%) were at least 10 calories under, 20 were within 10 either way, and 50 (19%) were at least 100 calories over (JAMA 2011). One side dish was over by 1,000 calories.

Their earlier paper picked the lowest-calorie items on each menu. Those 29 restaurant foods measured 18% above the stated value, and 10 supermarket frozen meals measured 8% above (J Am Diet Assoc 2010).

Packaged food has a legal tolerance, and it is wide. A food is misbranded only if the composite is "greater than 20 percent in excess of the value" declared (21 CFR 101.9(g)(5)). The same rule calls reasonable deficiencies acceptable, so there is no stated floor at all.

Do photo apps fix any of this?

No. The one controlled test I can find had four of them 250 to 345 calories low per meal (NIDDK, 2026).

Researchers at the NIDDK prepared 102 meals in a metabolic kitchen, weighed every ingredient to 0.1 g, photographed them, and ran the photos through MyFitnessPal, Lose It!, Cal AI and Appediet. The apps came in 250 to 345 calories under per meal and about 30 g under on fat (American Society for Nutrition, July 2026).

It is a conference abstract, so treat it as preliminary. I pulled the per-app breakdown apart in the Cal AI accuracy post, where the worst performer was also the best rated.

What do you do with a 330 calorie error?

Use the tracker as a consistent ruler, not as a scale. Set your target from your own logged average and your weight trend, and stop comparing it to a formula.

If you log 1,800 and the truth is 2,100, and you lose weight on that, then 1,800 in your app is your maintenance minus a deficit. The absolute number never needed to be right. What has to be stable is how you log, which is why the spread in those limits of agreement matters more than the bias.

Four things that actually shrink the gap, each one from a study above:

1. Log the meal as you eat it. Foster's single recall was 25% under, and a recall is what you do when you log at bedtime.
2. Log the forgotten things first. Boushey's underreporters missed 1,000 calories a day, and the items people photographed outside their provided meals were sugar sweetened beverages, coffee, tea and alcohol.
3. Distrust the menu before you distrust the app. The calorimeter found 50 of 269 restaurant items over by 100 calories or more.
4. Keep one logging habit rather than one accurate day. The limits of agreement are wide, so your average over a fortnight carries more information than any single total.

MealSaid is my attempt at the first of those. You say what you ate, it writes calories and macros to Apple Health, and it calls your phone at times you pick if you have not logged.

It has no barcode scanner, no photo logging, no food database to search and no Android build. It is English only and needs iOS 26. None of the portion-estimation problem above goes away, because you are still the one guessing how much rice was on the plate.

If you want a curated database and a long micronutrient list, Cronometer does that better than anything I have built. Why people stop logging at all is a separate problem, covered in the dropout research post.

How I checked this

Checked 23 October 2026. I read each paper at PubMed Central and took the numbers from the results text or the tables, not from abstracts alone or from secondary coverage. Sample sizes: 45, 30, 98 and 50 adults for the four validation studies, and 269 food items from 42 restaurants for the calorimetry work.

The 26 calorie database gap and the 13 to 1 ratio are my arithmetic on Evenepoel's Table 1 and Serra's Table 5, not figures either paper prints. The NIDDK meal study is a NUTRITION 2026 conference abstract and has not been peer reviewed, so it sits apart from the rest.

What I could not verify: any controlled comparison of two consumer calorie apps against the same weighed meals with the database held constant, which is the test that would settle the database question properly. The eCFR site refuses scripted fetches, so the regulation link goes to the Cornell copy of 21 CFR 101.9. Corrections to ohmbrock42@gmail.com.

Those two numbers only agree if the average meal contains 1,725 calories. The meals in that study were ordinary plated meals from a metabolic kitchen.

I read the NIH press release, the trade coverage carrying the per-app figures, and Cal AI's own site and App Store listing on 22 October 2026. I build a calorie tracker myself, so the last section says where mine is no better.

How accurate is Cal AI?

It underestimated meals by 345 calories on average, with a 95% confidence interval of 296 to 392, across 102 controlled meals.

Olivia Charles and Aaron Hengist of the NIH's NIDDK intramural programme worked from 102 standardised meals made in the metabolic kitchen at the NIH Clinical Center, every ingredient weighed to 0.1 g. They photographed each meal, ran the photos through four apps, and compared the output with food composition tables (American Society for Nutrition).

Every app came in low. The headline was an underestimate of roughly a third of the energy on the plate.

Caveats before anyone quotes this at me. It is a conference abstract presented at NUTRITION 2026 and has not been through journal peer review (Healio). And the apps were fed photographs with no correction, which is how most people use them but not the best case.

One more thing the chart will not tell you. The intervals for Cal AI, Lose It! and MyFitnessPal overlap heavily, so this test does not establish a ranking among those three. Only the gap to Appediet at 252 comes close to separating.

Still, 102 weighed meals is a better test than any of these companies has published about itself.

Does Cal AI's own accuracy claim hold up?

Not against this test. "About 80% accurate" and a 345 calorie miss only reconcile at meals of 1,725 calories.

The claim sits in the FAQ on Cal AI's own site, unqualified (Cal AI). There is no sample size, no method and no link to data behind it.

Do the arithmetic the other way. If 345 calories is the mean error and the claim of a 20% error holds, the mean meal has to be 1,725 calories. A plated lunch is not that.

The ratings do not help you either. Cal AI holds 4.80 stars from about 372,000 US ratings, the highest of the four apps tested, while posting the largest error (App Store).

Why do photo apps miss so much fat?

Because fat is invisible in a photograph. All four apps were about 30 g of fat short per meal.

Hengist put it plainly: "These apps tend to underestimate calories, especially from fats" (American Society for Nutrition). Oil absorbed into a stir fry, butter melted into a sauce, ghee stirred through dal, dressing already tossed into a salad: none of it is visible once the plate is made.

Fat carries more energy per gram than protein or carbohydrate, so missing 30 g of it costs more than missing 30 g of anything else. Carbohydrates, the one macro you can mostly see on a plate, were estimated more consistently across all four apps (American Society for Nutrition).

Cal AI's pitch leans on hardware, saying the phone's depth sensor calculates food volume (Cal AI). Depth solves volume. It does not tell you what was poured into the pan twenty minutes earlier.

What does Cal AI cost?

$29.99 is the highest of five prices Apple lists for its Unlimited purchase, and Cal AI publishes no price page at all.

The In-App Purchases panel on the US listing shows "Unlimited" at $2.99, $5.99, $9.99, $19.99 and $29.99, plus a $0.99 Streak Restore, which buys back a logging streak you broke (App Store). Apple never prints the billing period, so none of those numbers tells you what you are committing to.

Cal AI's site offers a 3 day trial and states no price (Cal AI). Its terms mention auto-renewal without naming a period or an amount (Cal AI terms).

There is also no free tier. The App Store description carries one line in capitals at the end, that food scanning analysis results require a subscription. I went through the same panel for nine apps in what is actually free in a free calorie counter, and Cal AI was the only one whose core function is gated outright.

Is Cal AI worth it?

Not if you are buying accuracy. If you want the lowest effort way to get a number, it works, as long as you treat the number as a floor.

Here is the practical reading of the NIH result. If the pattern held across three meals a day, you would be short about 1,035 calories a day, which is more than most people's entire intended deficit.

That is a projection from a single-meal mean, not a measured daily figure, so treat it as the direction rather than the size. Even halved, it is enough to make a diary useless for the thing diaries are for.

If you want fewer calories missed, weigh your food and log it by weight. Cronometer does this better than anything I have built: lab-analysed nutrition data, a free barcode scanner and up to 95 nutrients on the free tier (Cronometer).

MealSaid, which I built, does not take photos at all. You say what you ate and it writes calories and macros to Apple Health (how MealSaid works), which moves the guess from the camera to your own description of the portion. That is a different error, not an absent one.

I have not published accuracy numbers of my own, so nothing on this page says MealSaid beats Cal AI. That test is running: weighed meals against a kitchen scale and USDA reference values, and it will go up with the raw table whichever way it falls. The gaps I can state now are that there is no barcode scanner, no photo logging and no food database search, it needs iOS 26 on an iPhone, and it is English only.

How I checked this

Checked 22 October 2026. The study facts come from the American Society for Nutrition release on the NIDDK work presented at NUTRITION 2026, and the per-app calorie figures and confidence intervals come from Healio's report of the same abstract, which is the only source I found carrying the breakdown by app. The abstract itself does not appear to be published anywhere I can link to.

The prices, the purchase names and the ratings were read directly off the US App Store listings on 22 October 2026 for all four apps. Cal AI's accuracy claim, the depth sensor line and the 3 day trial came from calai.app, and the auto-renewal line from its terms page.

What I could not verify: the average energy of the 102 meals, which no source states, so the "about a third" figure stays as reported rather than recomputed per app. Healio prints all four intervals with the larger number first, which fits differences expressed as negatives; I have shown them ascending. The 1,725 calorie and 1,035 calorie figures are my own arithmetic on the published means, with no new data behind them. Corrections to ohmbrock42@gmail.com.

On 22 October 2026 I opened the In-App Purchases panel on nine US App Store listings and wrote down every purchase Apple shows, then read each company's own page on what the free version keeps.

If you want a free calorie counter without a subscription hanging over it, the shortlist is short. Cronometer and MyNetDiary hand you a working tracker for nothing, Lose It! does not, and two of the nine have no free tier at all.

I built MealSaid, which is free and has no purchases, so take my last section with the salt it deserves. The table is from their pages.

What does free actually mean on the App Store?

It means free to install. All nine apps I checked sell a subscription inside, and the listed prices span $0.99 to $119.99.

Apple shows an In-App Purchases panel on every listing, up to ten entries long. Eight of the nine apps fill all ten slots. FatSecret shows nine.

Ten is a display cap, not a count. An app selling twice that many looks the same on the page as one selling ten.

Which free tiers actually track food?

Cronometer and MyNetDiary. Both keep the barcode scanner and the whole database free, which is the part everyone else charges for.

Cronometer's free features page is specific: "Unlimited use of our Barcode Scanner is included in a Basic account", along with the million-entry database and up to 95 nutrients (Cronometer). Gold buys you no ads and long-term charts (Cronometer support).

MyNetDiary goes further on paper. Its free tier page claims the barcode scanner, voice logging, 13 charts, reports across your full history and 108 nutrients, with no ads (MyNetDiary).

Lose It! is the surprise. Its App Store listing names three free basic features: calorie tracking, a weight loss plan and community groups. It puts the barcode scanner, photo logging, voice logging and macronutrients in Premium (App Store). A calorie counter whose free tier does not break out protein is a narrower product than the download button suggests.

MacroFactor is at least blunt. Its listing says: "MacroFactor has a free trial, but does not offer a free subscription tier" (App Store). Cal AI's listing carries one line in capitals at the bottom, that scanning results require a subscription, and its site offers a 3 day trial (Cal AI). Its cheapest listed purchase, $0.99, is a Streak Restore, which buys back a logging streak you broke.

Why does one app show seven different prices?

Because the panel lists what people buy, and the bigger apps run several price points at the same time. Lose It! shows seven of them under one name.

Seven separate purchases on the Lose It! listing are named exactly "Lose It! Premium", at $9.99, $11.99, $19.99, $23.99, $29.99, $39.99 and $79.99 (App Store). Its own help centre quotes one number: "$79.99 USD / Year" (Lose It! help).

The pattern repeats. MyFitnessPal shows Yearly Premium at $79.99 and at $49.99, Lifesum shows a one-year subscription at $119.99 and at $99.99, and Yazio shows 12 Months at $23.90 and at $47.90 (App Store).

The price you are shown in the app is one of several, and the renewal is the one that matters. The Lose It! help centre also sells a Lifetime unlock at $299.99, or $229.99 if you already pay for Premium (Lose It! help). I cannot reconcile that with the two Premium Lifetime purchases on the listing at $49.99 and $59.99.

Which free calorie counters show you ads?

Four of the nine declare advertising to Apple: MyFitnessPal, Lose It!, Cronometer and Yazio. The other five do not.

The declaration sits in the age rating panel of each listing, under Contains. It is what the developer told Apple rather than a test I ran, so treat it as a minimum.

MyFitnessPal is the only one that explains the trade in writing. The core of the app is free, it says, and "it stays that way because ads help pay for it" (MyFitnessPal support). Premium removes banners, sponsored posts and ads in the forums.

Cronometer sells the same thing, listing "No Advertisements" first among Gold benefits (Cronometer support). So on the two best free tiers here, one of them is funded by ads and the other says it is not.

So which one should you install?

If you want a free tracker that will still work in a year, install Cronometer or MyNetDiary. Everything else on this page is a trial with a database attached.

Pick by what you log. Packaged food means you need the barcode scanner, which rules out the free tiers of MyFitnessPal and Lose It!. Micronutrients mean Cronometer. No ads and free means MyNetDiary, on its own account.

If you are leaving MyFitnessPal specifically, the paywall history and the full price comparison are in our MyFitnessPal alternatives post.

MealSaid, which I built, is free with no ads and no in-app purchases. That is the only claim on this page I cannot show you an App Store panel for, because the listing is not public yet.

You say what you ate and it writes calories and macros to Apple Health, and at times you pick it calls your phone and asks. It has no barcode scanner, no photo logging, no database to search and no way to import food other apps wrote to Apple Health. It needs iOS 26 on an iPhone and it is English only. How MealSaid works has the detail.

Free with nothing to buy is easy when an app does less. Cronometer gives you 95 nutrients for nothing, which is more generous than anything I have built.

How I checked this

Checked 22 October 2026. I fetched the nine US App Store listings and read the In-App Purchases panel on each one, recording the purchase name and price exactly as Apple shows them, plus the advertising flag in the age rating panel. Counts are 10 for every app except FatSecret at 9, and ten is Apple's cap.

Free tier contents come from each company's own page: Cronometer's free features page and Subscription Types article, MyNetDiary's free tier page, the MyFitnessPal plan comparison, and the App Store descriptions for Lose It!, MacroFactor, FatSecret, Lifesum, Yazio and Cal AI. The Lose It! lifetime prices are from its help centre.

What I could not verify: FatSecret's own premium page returns 403 to anything scripted, so its free and paid split here comes from the App Store description instead. Lifesum and Yazio never name which features the free tier loses, so those two cells say so rather than guess.

A third-party pricing blog puts Yazio's barcode scanner behind Pro, and I am not repeating a split Yazio itself does not publish. I installed none of these apps for this post, so every ad claim is a declaration rather than an observation. Corrections to ohmbrock42@gmail.com.

I read each app's own help pages and US App Store listing on 21 October 2026 rather than trusting the marketing word "voice". MealSaid is my app, so treat the other six rows as careful docs reading and not a hands-on test.

What does voice input mean in a calorie app?

Three levels, and apps use the same word for all of them. Level one is the iOS Dictate key, level two is a microphone that searches for one food, level three is a spoken sentence that becomes a set of entries.

Level one costs nothing and belongs to Apple, not the app. Any app with a text field gets it, because the keyboard has a Dictate key (Apple). You still search, and you still pick the match and the serving.

Level two is a microphone button wired to the same search. Lose It calls it Say It, and its own logging guide puts it plainly: "use Say It to search for a food by voice" (Lose It).

Level three is the one people picture when they imagine talking to a tracker. You say the sentence you would say to a friend, and the app hands back several foods at once.

Which calorie apps let you speak a whole meal?

Six that I could confirm from their own documentation: MyFitnessPal, Cronometer, MyNetDiary, MacroFactor, Lifesum and MealSaid. Lose It has voice, but only the search kind.

MyFitnessPal is the clearest writer of the group. Its help page says "Voice Logging searches the same food database you'd search by typing", then lists what is not supported yet: saved foods, recipes, meals, Quick Add, weight, exercise and water (MyFitnessPal).

MacroFactor's guide is equally blunt about the mechanism: "Just type (or speak) a description of your food or meal" (MacroFactor). The speaking part is just the keyboard; the matching is the feature.

What does spoken meal logging cost a year?

Between $59.99 and $99.99 a year in every app except mine, with a median of $79.99. Cronometer Gold is the cheapest paid route at $59.99 a year (Cronometer).

MyNetDiary is the interesting case, because it sells both levels. Its free tier page lists voice logging among the free features, while its App Store listing puts "AI voice logging: describe your meal out loud" under Premium Plus (App Store).

Premium Plus is $99.99 a year against $59.99 for plain Premium (MyNetDiary). A company with no reason to draw the distinction has drawn it, in its own price list, which is why I trust it more than any marketing page here.

MacroFactor has no free tier at all, so its $71.99 a year buys the app as well as the Describe feature (App Store). Lifesum's cheapest annual SKU is $99.99, with a $119.99 one also listed (App Store).

Can I log by voice without opening the app?

Only Cronometer and MealSaid document it, and both do it through Siri. Cronometer's support article says the feature "allows you to log foods using Siri without having to open the app", for Gold users on iOS (Cronometer).

Everyone else puts the microphone behind a screen. MyFitnessPal's numbered steps begin with opening the app and tapping the plus icon, Lifesum's AI flow opens from the plus button in the Diary, and MacroFactor's Describe lives inside the food logger.

That gap matters, because the tap people skip is the first one. I walked the routes on my own phone in the post on logging without opening an app.

Cronometer does one thing here that I like: it tells you to review Siri-logged foods the next time you open the app, and marks them with a Siri icon. Honest about the failure mode.

Is plain dictation good enough on its own?

For one food, yes, and it is free in every app. For a plate with five things on it, you are doing five searches with your voice instead of your thumbs.

Apple's guide warns that "Dictation may not be available in all languages or in all regions" (Apple). On supported models it runs without an internet connection, which no server-side voice feature can match.

Which one should I pick?

If you already pay for a tracker, use the voice feature you have bought; if you are choosing from scratch, pick on what happens after you speak. The question that sorts the field: does it take a whole sentence, or one food at a time?

Then, can you start it without opening the app? Whether the app shows you its matched foods before saving them only starts to matter once you log every day, and by then it matters a lot.

Cronometer passes all three, for $59.99 a year, and it has the deeper nutrient data too. If I were paying, that is where I would start, and our MyFitnessPal alternatives table has the rest of its free tier.

MealSaid is my answer to the same three tests, for free. You say what you ate, it estimates calories, protein, carbs, fat and fiber, and writes them to Apple Health; a Siri phrase does it with the app closed. It has no barcode scanner and no photo logging, there is no food database to browse, it is English only, and it needs iOS 26. Wrong app for anyone who logs packaged food by barcode.

How I checked this

Checked 21 October 2026. I read each company's own help article or US App Store listing on that date and took the prices from the listing or the pricing page, not from a roundup. MyFitnessPal's help article was last updated 21 September 2026 and Cronometer's voice-logging article on 20 October 2026, per their help centres.

Re-checked 26 October 2026: Lifesum's AI Tracking help link returned a 404, so I replaced it with Lifesum's current "AI Tracking - How does it work?" article, which still says voice, text and photo input run from the plus button and need Premium. Prices and other claims in this post were unchanged.

Apple's current iPhone guide page renders its body only with JavaScript, so the Dictate key link above points at the archived iOS 12 copy of the same article, which still loads as plain text.

Two apps are missing for a reason. FatSecret's help pages and premium page returned 403 to every fetch I tried, so its Smart Assistant is not in the table rather than guessed. Cal AI's own site offers describing a meal in words (calai.app) but neither that page nor its App Store listing mentions voice, so I left it out too.

I did not install and time these apps. The one thing I can say from use is about my own, and the timing test across six apps is still running. Corrections to ohmbrock42@gmail.com.

I read each app's own help centre, App Store listing or public feedback board on 21 October 2026. MealSaid is my app, and I have no adherence numbers of my own yet, so the last row is a description of the feature and not a claim that it works better.

Which calorie apps actually send a logging reminder?

MyFitnessPal, Cronometer and MyNetDiary send push notifications, MealSaid places a call, and I found nothing for Lose It or MacroFactor. The more useful question is whether the reminder is skipped once you have logged.

MyFitnessPal's setting is worded well. The row you tap reads "Remind me if I haven't logged", then you pick the meal and the time (MyFitnessPal).

Cronometer does the same thing and goes further, with a separate toggle per diary group so breakfast and dinner can nag independently (Cronometer).

MacroFactor is the surprise. Its public feedback board carries a Weigh-In Notification request at 52 votes, still marked Under Consideration (MacroFactor), and I found no food logging reminder documented at all.

What happens if you ignore a logging reminder?

Nothing, in all six. None of them escalates, and the research says most prompts are ignored anyway. The Lock Screen stacks them up and that is the end of it.

Chen and colleagues ran the cleanest test I could find. Thirty people used a weight maintenance app paired with FatSecret for four weeks, each person seeing both reminder styles, with reminders either at fixed times or timed to a change in their physical activity.

Read the left-hand dots first. Even the better style left four reminders in five untapped, and that is among people who had volunteered for a logging study and knew they were being watched.

Timing did help with speed. People clicked 12.33 minutes after a context-triggered reminder against 18.42 minutes for a fixed-time one (Chen et al., CHI 2024).

What it did not move was the thing you care about. Days with a log went from 55.54% to 58.87%, a difference of 3.33 percentage points with a 95% confidence interval of minus 3.07 to 9.73, which includes zero (Chen et al., CHI 2024).

The bigger study points the same way. Bidargaddi and colleagues micro-randomized push notifications across 89 days for 1,255 users of a wellbeing app and found a risk ratio of 1.039, with a 95% confidence interval of 1.01 to 1.08 (JMIR 2018). Real, and tiny.

Does a different kind of reminder work better?

I could not find a published test of anything louder than a push notification for food logging, and that includes my own app. What the literature does show is that the size of the push effect is small and that effort, not forgetting, is the bigger problem.

Bentley and Tollmar saw logging rise from 12% to 63% when notifications were added, but across two different studies rather than a controlled comparison (CHI 2013). I went through that and six other dropout studies in why people quit calorie tracking.

MealSaid takes the loud route: at times you pick, your iPhone rings like a normal call, a voice asks what you ate, and it waits while you answer. The call is skipped if you logged recently, the same conditional logic MyFitnessPal and Cronometer use for their banners.

I believe a ringing phone is harder to dismiss without deciding than a banner is. I have not proved it. A 30 day test of calls against reminders against a widget, same person, is running, and the raw table goes up whichever way it lands.

How do I turn reminders on?

Both apps that document it hide the setting a few taps from the home screen, and neither charges for it. Reminders are not in the main settings list, which is probably why people think the feature is missing.

In MyFitnessPal on iOS: tap the profile icon, scroll to Reminders, tap the plus, choose the meal, choose the time, tap the check mark. The help page adds the usual trap, that iOS notification permission has to be on or nothing arrives.

In Cronometer: More, then Account, then Notifications, then Allow Notifications, then Reminder to Log Food, then pick a time. Diary group reminders and a weight reminder after seven quiet days sit on the same screen.

If you use MyNetDiary, meal reminders are listed among its free features on its US App Store listing, though I could not find a support page with the steps (App Store).

What should you do if reminders are not working for you?

Stop adding more of them. Four reminders a day that you swipe away teach you to swipe, and the trial numbers above say the tap rate is low even when the timing is clever.

Two things are worth trying instead. Cut what a single entry costs you in taps, and put the log somewhere you already look, which is what Lose It's widgets are for.

MealSaid is my version of the first idea, and it is free with no in-app purchases. The gaps are real: no barcode scanner, no photo logging, no food database to browse, English only, and it needs iOS 26. If you log packaged food by barcode, MyFitnessPal or Cronometer will serve you better than I will.

How I checked this

Checked 21 October 2026. I read the MyFitnessPal reminders article (updated 1 October 2026) and the Cronometer notifications article (updated 20 October 2026) through their help centres, and searched Lose It's help centre for reminders, notifications and push notifications without finding a logging reminder article.

The two studies were read in full text. Chen et al. is 30 participants over four weeks, within subjects, so treat every percentage as a small sample. Bidargaddi et al. is 1,255 users over 89 days, which is large but measures any app interaction rather than food logging specifically.

Two gaps I could not close. MyNetDiary publishes no support page for its reminder settings that I could find, so its row rests on the App Store listing. MacroFactor could have an undocumented reminder in the app; its own feedback board suggests not. Corrections to ohmbrock42@gmail.com.

I read these papers because I was about to build another tracker and wanted to know what kills them. Every number links to its source with the sample size. Checked 20 October 2026.

How many people actually keep logging?

Almost nobody, past the first week. The best large-scale number is 2.58%; the best trial number is a fall from 97% to 35% of participants logging in between month one and month six.

The Eatery was a free iPhone app that logged meals by photo. Helander and colleagues had the data for every download between October 2011 and April 2012: 189,770 people used it at least once. "Active" meant ten or more photos and at least a week of use.

Result: "only 2.58% (4895/189,770) used the app actively". Another 11% were semi-active and 86% never took a valid photo (J Med Internet Res 2014).

The MyFitnessPal trial is the one people misquote, including me until this week. Laing and colleagues randomized 212 overweight primary-care patients, half to MyFitnessPal. Logins in the app group ran 97% in month one, 55% in month two, then 47%, 43%, 23% and 35% in month six; median logins went from 8 a month to zero. Weight loss was no different from usual care.

At six months 80% said they planned to keep using the app, a number I had inverted in my notes as "20% would continue". The honest reading is worse: most said they would continue, and most did not (Ann Intern Med 2014).

The reviews agree on the shape. Burke, Wang and Sevick's review of 22 self-monitoring studies describes "a gradual decline in self-monitoring adherence" wherever it was measured over time, with one study where only 25% were still logging at the end (J Am Diet Assoc 2011).

Goldstein and colleagues measured dietary logging in a coached 12-week program at 92.7%, 94.9% and 86.7% by month, and found it "decreased significantly over time for eating and weight" (Obes Sci Pract 2020). That is with coaching; the free-app curve above is without.

Industry numbers are rougher. Adjust's benchmark has Day 7 retention at 13% and Day 30 at 7% across all app categories, and health and fitness at 27% on Day 1 (Adjust). I could not trace the "health apps keep 12 to 16% at Day 7" figure that circulates in blog posts to a primary report, so I am not repeating it.

Why do people stop?

Effort first, forgetting second, social situations third, in every study that asked.

Cordeiro's paper is the one to read. It is qualitative, so the counts above are respondents, not percentages, but it names the mechanism: every entry costs a search, a portion guess and a decision, and people forget until the moment has passed.

"Simply forgetting to journal is a major barrier to reliable journals". Lapsers also felt judged by red numbers after a bad day and stopped opening the app, the "negative nudge" in the title.

Does making logging simpler actually help?

Yes, in the one randomized test I could find, and by a lot: 97% of days logged versus 49%, with the same weight loss.

Patel and colleagues ran a fully remote pilot with 38 adults over 84 days. Eighteen logged every food in the Fitbit app. Twenty used a daily checklist of "red zone" foods, no calorie counting.

Median days logged: 49% (interquartile range 5 to 67) detailed, 97% (86 to 100) simplified. Weight change 3.4 kg versus 3.3 kg, a wash (JMIR Form Res 2022). Thirty-eight people is a pilot, not a verdict, but it is the only head-to-head I know of and the gap is not subtle.

Do reminders fix forgetting?

Partly. Notifications roughly quintupled logging in one early study, and context-triggered prompts beat clock-based ones in a 2024 randomized trial. But that effect was 21.8% versus 17.3%, so most prompts are still ignored.

Bentley and Tollmar added daily notifications to a wellbeing-logging app and saw logging go from 12% in a 10-user pilot to 63% in a 60-user study (CHI 2013). Two different groups, not a controlled comparison, so read it as "reminders matter a lot", not "reminders give you 5x".

Chen and colleagues did it properly in 2024, randomizing fixed-time notifications against context-triggered ones. Context won, 21.77% to 17.32% of prompts followed by a log within the hour (CHI 2024). The "meal-time prompts improve logging 64%" figure from my notes is in neither paper; the 63% above is probably where it came from.

Both papers share one thing: a notification is still a request to open an app and do the work. It fixes forgetting and leaves effort alone.

What does this mean for picking a tracker?

Pick for the two failure modes the research found: minimise the effort of one entry, and have something that interrupts you before the meal is forgotten. Accuracy matters less than you think, because an entry that never happens has infinite error.

Four tests before you commit:

1. Time one real meal from Home Screen to saved. Over about 20 seconds and the Laing curve is coming for you.
2. Check what happens when you forget. Is there a reminder? Can you log yesterday's dinner in one step?
3. Check whether a bad day punishes you. Cordeiro's lapsers stopped opening the app after red numbers.
4. Check that meals land in Apple Health, so switching apps later keeps your history. Who writes and reads is in the Apple Health post.

MealSaid is my attempt at the first two, and it is one answer, not the only one. You say what you ate, by voice or by a Siri phrase with the app closed, so an entry is one sentence instead of a search per ingredient. At the times you pick, it calls your phone like a real call and asks what you ate; a call is harder to swipe away than a banner, which is the whole idea.

The caveats:

• no barcode scanner
• English only
• iOS 26 required
• I have not yet published my own adherence numbers

That 30-day test (reminders versus widget versus scheduled call, same person, days logged per condition) is running and will go up with the raw table whether or not the call wins. Patel's checklist, or a photo-first app, are reasonable answers for people who do not want to talk to their phone.

How I checked this

Checked 20 October 2026. Each study was opened at the linked publisher or PubMed Central page and the numbers read from the full text or abstract, not from secondary blogs. Two figures from my earlier notes were wrong and are corrected above: the MyFitnessPal "20% would continue" (it is 80% said they would, and 35% logged in during month six), and "meal-time prompts plus 64%" (no such number; Bentley and Tollmar report 12% to 63%).

I also dropped a Day 7 and Day 30 health-app retention range I could not trace to a primary report. ACM Digital Library pages return 403 to scripts, so the Cordeiro and Bentley links go to open copies; the DOIs are 10.1145/2702123.2702155 and 10.1145/2470654.2466140. Corrections to ohmbrock42@gmail.com.

Every price and feature below was checked against the company's own page or its US App Store listing on 20 October 2026. I built MealSaid, so it is last in the table, with where it loses.

What did MyFitnessPal move behind the paywall?

Barcode scanning on 1 October 2022, Voice Log in December 2024, and Meal Scan (photo). The free tier keeps manual logging, the database and Apple Health sync, with ads.

The three moves, each from MyFitnessPal's own pages:

• Barcode scanning. MyFitnessPal's own help page is blunt about the barcode change: "As of October 1, 2022, Barcode Scan is available only with a Premium subscription" (MFP support). The change was announced on 24 August 2022 (Droid Life).
• Voice Log followed in December 2024, and the help page says "Voice Logging is currently available for Premium users on iOS and Android" (MFP support).
• Meal Scan is listed under Premium in the Free, Premium and Premium+ comparison (MFP support).

Ads: "The core of MyFitnessPal is free, and it stays that way because ads help pay for it" (MFP support).

Still free, and credit where due: the biggest food database in the category, search logging, recipes, and writing meals to Apple Health with no Premium gate (MFP support). If you log packaged food by barcode, Premium is still a reasonable buy. If you log by search anyway, free is fine apart from the ads.

What does MyFitnessPal Premium cost in 2026?

Premium is $19.99 a month or $79.99 a year. Premium+ (adds meal planning) is $24.99 a month or $99.99 a year.

Both are from MFP's own pricing post (MFP blog) and match the in-app purchase list on the US App Store (App Store), which also shows $49.99 and $9.99 promotional or legacy SKUs.

Which alternatives are worth switching to?

Seven, for different reasons: Cronometer and MyNetDiary for free barcode scanning, MyNetDiary and MacroFactor for no ads, Lose It for a like-for-like swap, Cal AI for photos, FatSecret for a quiet free tier, MealSaid for voice.

Who is each one for?

Pick by what you log and how you feel about ads.

Cronometer is the pick if you want the most for free. Its basic account is "for an unlimited amount of time" (Cronometer support) and includes "Unlimited use of our Barcode Scanner" (Cronometer). Ads until you pay $59.99 a year for Gold. Smaller database than MFP, far better curated for whole foods, and it tracks micronutrients MFP does not.

MyNetDiary is the pick if ads drove you out. Its free tier promises "no ads, no account required to start, no expiration, and no scanner paywall" (MyNetDiary), and plain voice logging is free. Premium is $59.99 a year. The interface is dated; that is the trade.

Lose It is the closest like-for-like swap. Its About page lists barcode scanning in the free version (Lose It), though its tiers page puts "barcode scanning, smart photo and voice logging features" under Premium (Lose It); I could not resolve that from their docs, so check on a fresh install.

Premium lists at $79.99 a year but Lose It says it "adjusts prices based on various factors like seasonal promotions" (Lose It), so the $39.99 you see quoted elsewhere is a promo, not the list price. Ads on free, with a one-time purchase to remove them.

MacroFactor is the pick if you are willing to pay and hate ads. "There isn't (and will never be) a free version" and "We will never serve ads or use ad networks to track you" (MacroFactor). $71.99 a year. Its adaptive calorie targets and its two-way Apple Health sync are the best in the group. It is the app I would use if I logged by barcode and weighed food.

Cal AI is the pick if you want to point a camera at a plate and nothing else. The photo scan is subscription-only ("FOOD SCANNING ANALYSIS RESULTS REQUIRE A SUBSCRIPTION", App Store) after a 3-day trial.

Two cautions: Cal AI publishes no price page and no Apple Health write documentation that I could find, and an NIH study presented at NUTRITION 2026 found four photo apps, Cal AI among them, under-estimated meal energy by about 33% across 102 controlled meals (Healio).

FatSecret is the quiet option. Free search and barcode, an old community, and App Store SKUs of $59.99 a year for photo and voice. FatSecret's own premium page gives no US price (FatSecret) and I could not find an official statement on ads, so that column is blank rather than guessed.

Where does MealSaid fit, and where does it not?

MealSaid fits if your problem with MFP was the number of taps, and it does not fit if your problem was the barcode paywall, because we have no barcode scanner at all.

You say what you ate, by holding a mic or by "Hey Siri, log in MealSaid, two eggs and toast" with the app closed, and calories, protein, carbs, fat and fiber are written to Apple Health in about a second. If you forget, the app calls your phone at times you picked and waits while you list the day. Free, no ads, no in-app purchases.

What it does not do:

• no barcode, no photo
• no database you can browse
• no micronutrients beyond fiber
• no Android, English only
• it requires iOS 26

The estimates come from a language model checked against USDA values, so they are estimates; a weighed plate in Cronometer will be more precise. If you need any of those things, pick from the table above and use MealSaid for the meals you would otherwise skip, since everyone on this list writes to the same Apple Health ledger (who writes and who reads).

How I checked this

Checked 20 October 2026. For each app I opened the US App Store listing and read the in-app purchase list, then opened the company's own pricing or help page and the Apple Health page where one exists. Prices are US dollars as shown that day; several companies say they vary by region and promotion, and I have flagged those.

MyFitnessPal, Lose It, Cronometer and FatSecret help pages return 403 to scripts, so they were read in Safari. The MacroFactor App Store ID in my earlier notes was wrong and 404s; the live listing is id1553503471. I have not timed or measured accuracy for any of these apps in this post; that is the next two posts. If a price has changed by the time you read this, email ohmbrock42@gmail.com and I will update the table with a new dated row.

Checked 20 October 2026 on my iPhone running iOS 26. I built MealSaid, so read the competitor section as "what I could find in their docs", not a hands-on test of each one.

Why does it matter whether the app opens?

Because opening a tracker is the step people skip. Every tap between "I just ate" and "it is logged" is a chance to decide you will do it later, and later is usually never. A stopwatch comparison across six apps is coming as its own post; this one is about the routes.

How do I log with Siri and MealSaid?

Say "Hey Siri, log in MealSaid, two eggs and toast" and let go. Siri shows a card confirming what was logged. The app does not open.

The phrase has three parts, and only the middle one is fixed:

1. "Hey Siri" (or hold the side button).
2. "log in MealSaid" or "tell MealSaid". Apple requires the app name in third-party Siri phrases; its own guide says you run a shortcut by saying its name (Apple Support). Not our rule.
3. What you ate, in plain words. "Two eggs and toast." "Chicken burrito from Chipotle, no rice." "Did chest for 45 minutes" logs a workout instead.

Your words go to our server, a language model turns them into entries with calories, protein, carbs, fat and fiber, and those are written to Apple Health. About a second. Say when ("this morning at 8") and it is logged at that time.

Two honest limits. The Siri route needs an internet connection, because the understanding happens on our server, not on the phone. And it is English only.

How do I set up the Lock Screen control?

Press and hold your Lock Screen, tap Customize, tap Lock Screen, tap the minus on the flashlight or camera control, then tap the plus and choose MealSaid. From then on it is one tap from the locked phone to a listening mic.

The steps as they appear on my phone, iOS 26:

1. Wake the phone, do not unlock it. Press and hold on the Lock Screen until the gallery appears.
2. Tap Customize under your current Lock Screen, then Lock Screen (not Home Screen).
3. At the bottom you see two controls, flashlight and camera by default. Tap the minus on the one you use less.
4. Tap the plus that replaces it. Scroll or search for MealSaid and pick the control.
5. Tap Done, top right.

Apple documents the flow under Create a custom Lock Screen. Third-party controls exist since iOS 18, so any app that ships one can live here.

Press and hold the control, the same firm press the flashlight needs. The phone unlocks with Face ID, the app opens straight into the listening screen, and you talk.

The MealSaid control opens the app rather than recording from the Lock Screen, so this route is one press plus speaking, with no searching.

How do I add the Home Screen widget?

Press and hold an empty spot on the Home Screen, tap Edit, tap Add Widget, search MealSaid, and add the medium one. Tapping it opens the app already listening and shows today's calories against your target.

Apple's steps are at Add, edit, and remove widgets. The MealSaid widget shows a mic button, today's total and a progress bar. Tap anywhere on it.

Which route I use depends on where I am:

• The widget is my route at home, where the phone is unlocked on the table.
• Siri is for cooking or driving.
• The Lock Screen control is for the gym.

Which route is fastest?

Siri, because nothing opens and there is nothing to tap. The other two routes cost one tap plus whatever you say.

What can Shortcuts and Siri do with other calorie apps?

Less than you would hope, and the useful bits are mostly paid. The pattern across the big apps is: voice logging exists, but as a button inside the app, behind the subscription.

App by app, from their own pages:

• MyFitnessPal's announcement calls Voice Log "available to Premium members" and says you reach it from the app's navigation bar, so the app is open before you speak.
• Lose It's listing puts "Say It!" in its Premium plan.
• MacrosFirst is the most Shortcuts-friendly: its help page says to "say 'Hey Siri…' followed by one of the phrases below", and water and weight finish inside Siri; food is Premium and may hand off to the app.
• Cronometer is the closest to our route: its doc says Gold users can "log foods using Siri without having to open the app", by saying "Log food in Cronometer" and then answering Siri's question about food, serving size and diary group. Free users do not get it.

Can I build my own Siri shortcut for an app that has none?

You can build a shortcut that opens the app and dictates text, but you cannot make it save an entry unless the app exposes a "log food" action to Shortcuts. Most do not.

In the Shortcuts app, tap Add Action and look under the app's name. If it lists a "Log Food" action with a text field, chain Dictate Text into it and name the shortcut something you can say. If it only lists "Open MyFitnessPal", opening it is all Siri can do.

The fallback for any app is Apple's keyboard dictation: open the search screen, tap the mic on the keyboard, say the food, pick the match by hand. Not hands-free, but no typing.

Where MealSaid does not fit

MealSaid is for people who log by talking; if you weigh everything, pick a database app.

• No barcode scanner, no photo logging and no food database to browse.
• It estimates from what you said, checked against USDA values for common foods, and the estimate can be off; tap the entry to fix a number.
• It needs iOS 26, is English only, and has no Android version.

If you weigh everything and want gram-level control, Cronometer or MacroFactor will serve you better.

How I checked this

Checked 20 October 2026. Siri, Lock Screen and widget steps were done on my iPhone on iOS 26 with MealSaid 1.0.1 from the App Store, following each route from a locked phone and confirming the entry in Health > Browse > Nutrition. Competitor facts come from the linked pages (PR Newswire release of 11 December 2024, Lose It App Store listing, MacrosFirst and Cronometer help centers; Cronometer's article was last updated about three months before I read it), all opened the same day.

I did not install the competitors for this post; the timed comparison is the next one, with a CSV. If a link above says something different by the time you read it, email me at ohmbrock42@gmail.com and I will correct it.

MealSaid is my attempt at the opposite. You say what you ate. It is logged. That is the whole app.

This post explains what actually happens when you use it, so you can decide if it is for you.

What happens when you log a meal

You speak, the words become entries with calories and macros, and they are written to Apple Health in about a second.

You hold the mic button and say something like "two eggs, toast and a coffee" or "did chest for 45 minutes". Then you let go.

1. Your iPhone turns your speech into text. For the mic button this happens on the device, using Apple's speech recognition.
2. The text goes to our server, which asks a language model (Anthropic's Claude) to turn it into entries: a food or a workout, an amount, and an estimate for calories, protein, carbs, fat and fiber. For common foods it checks against USDA FoodData Central values.
3. The entries come back to your phone and are written to Apple Health as dietary energy, protein, carbohydrates, total fat and fiber. Workouts are written as real workouts with active calories.

Start to finish, this takes about a second. If you said when it happened ("this morning at 8"), it is logged at that time.

Tap any entry and you see the exact words it was estimated from, and you can fix a number if the estimate is off.

There is no search box, no barcode scanner and no portion slider. There is a keyboard fallback for when the gym is too loud, but I do not expect you to use it.

You do not have to open the app

Siri, the Lock Screen control and the Home Screen widget each get you to a logged meal without finding the app.

Three shortcuts cover most of my own logging (step-by-step setup for each is here):

The app calls you

At times you pick, your iPhone rings with a normal call from "MealSaid", a voice asks what you ate, and it waits until you are finished.

This is the part that is unusual, so let me be precise about it.

In the Profile tab you pick the times you usually forget to log, say 9 pm. At that time our server sends a push to your phone and your iPhone rings with a normal incoming call screen from "MealSaid". You can answer it from the Lock Screen like any call.

When you answer, a voice asks what you ate today and whether you worked out. Then it waits. It does not cut you off after three seconds the way a voice assistant does. You can pause, think, list everything from breakfast to the snack you are not proud of, and it only replies when you are actually finished.

If you correct yourself ("actually that was two slices, not one") it fixes the entry. When you say "that's all", it hangs up and the entries appear on your Home screen and in Apple Health.

Some details, because they matter:

• Calls only ring at the times you picked, or when you ask. You can say "call me" into the mic at any time and it rings you back in about five seconds.
• If you already logged recently, the scheduled call is skipped.
• It never dials a phone number or another person. It is a two-way voice conversation between you and the app, carried over your internet connection.
• The voice on the call is OpenAI's Realtime model, streamed through our server. Nothing on the call is recorded.

Seeing your day

Three places show your numbers: the Home screen for today, the Progress tab for the last 7, 30 or 90 days, and a widget.

• The Home screen shows today's calories and macros against your targets, with a week strip you can scrub.
• The Progress tab shows calorie bars per day for the last 7, 30 or 90 days, plus protein per day and your streak.
• There is a Home Screen widget with today's numbers.

Your targets come from onboarding, which is also a conversation: you say your weight, height, goal and how you train, and the app sets daily calorie and macro targets. Change your mind later and just say so.

Where your data lives

Your log lives on your phone, in Apple Health and in your account on our server; your voice is processed and discarded; Health data stays on the phone.

I want to be plain about this because "voice" and "AI" make people nervous, reasonably.

• Your log is stored on your iPhone, in Apple Health, and in your MealSaid account on our server (hosted on Fly.io), so a new phone keeps your history.
• Sign in is once, with Sign in with Apple. "Hide My Email" works fine.
• Apple Health data never leaves your phone. The app reads your workouts, steps, weight, heart rate and sleep on the device to give the call context, and writes meals and workouts there. None of that is uploaded.
• Your voice is processed, not kept. Speech is streamed to our server and on to OpenAI (the voice on the call) and Anthropic (understanding what you said), then discarded. Audio is not stored. The text of what you said is saved so you can read it under each entry.
• No ads, no tracking, no analytics SDKs. We count which features are used, with no food names attached, and that is it.
• Delete account in the Profile tab removes your account and every row about you from our server and wipes the app on your phone. Entries already written to Apple Health stay there until you delete them in Health.

The full detail is in the privacy policy.

What it is not

Not a scale, not medical advice, and not an offline app.

MealSaid gives you calorie and macro estimates from what you said. They are estimates, made by a language model from your description and public nutrition data, and they can be off.

It also needs an internet connection to log. Reading your log and your progress works offline, but understanding what you said does not happen on the phone.

Requirements

• An iPhone running iOS 26 or later, a microphone and an internet connection.
• Apple Health is optional, but it is the point.
• English only for now.
• Free, with no in-app purchases.

If you want to try it, the links are below. If something does not work, email me at ohmbrock42@gmail.com and I will answer within two working days.

MyFitnessPal, Cronometer, MyNetDiary, FatSecret, Lifesum, Carbon Diet Coach and MealSaid all write and do not read.

That one-way street is what makes "syncs with Apple Health" such a slippery phrase. Here is what each app actually does, with the doc it comes from.

What does "syncs with Apple Health" really mean?

It can mean three different things and app listings rarely say which. Writing nutrition into Health, reading nutrition out of Health, or reading other data like steps and weight.

Apple Health is a database on your phone. Apps ask permission to write to specific fields and to read specific fields, one by one. Apple publishes the nutrition fields it supports, 39 of them, from dietaryEnergyConsumed and dietaryProtein down to individual vitamins (Apple developer docs: nutrition type identifiers).

You can see exactly what each app asked for under Settings > Health > Data Access & Devices (Apple Support: share your Health data).

So when an app says it "syncs", ask three questions.

• Does it write my meals in?
• Does it read meals other apps wrote?
• Does it only read steps and weight and call that sync?

Which apps write nutrition and which read it?

Seven of nine write only. Two read. Nobody is a complete two-way mirror.

Two rows deserve a note. MacroFactor's import has a priority rule: manual entry beats its own food log, which beats Apple Health, and a MacroFactor food entry for a day blocks the Health import for that day (MacroFactor help).

Lose It marks imported dietary energy with a heart icon in the log so you can tell it apart; its doc says "Food data from other 3rd party apps in Apple Health may appear in Lose It!"

Why does it matter whether an app reads nutrition back?

Because reading is what lets you switch apps, or use two, without losing the day. A write-only app is a one-way door: its entries land in Health for everyone else to see, but it never looks at what anyone else logged.

In practice that means if you log breakfast with a Siri phrase in one app and lunch by barcode in another, only Apple Health itself has the whole day. MyFitnessPal will show you its own meals. Cronometer will show you its own.

Health > Browse > Nutrition > Dietary Energy shows both, which is why I tell people to treat the Health app as the ledger and the trackers as input devices.

MacroFactor is the one app here built for that two-app life, and it is the only one whose doc explains the conflict rules. Lose It gets you partway (energy only, and no macros unless you pay and set a custom goal).

Why are my calories doubled in Apple Health?

Two sources wrote the same meal, and Health adds them up. Fix it under Health > Browse > Nutrition > Dietary Energy > Data Sources & Access by turning off the source you do not want counted.

This is the trap the "sync" word hides. Every write-only app above will happily add its entries next to another app's entries. MyNetDiary's own support page says it directly: if you log the same food in two apps, "you may see duplicate records in Apple Health".

It is not only two apps either; the Athlytic recovery app's help doc notes that "MyFitnessPal has been known to double-log calories in Apple Health" on its own (Athlytic help, updated 24 Feb 2025).

A reader-plus-writer is a different risk. If an app reads dietary energy from Health and also writes dietary energy to Health, a second tracker's entries can get imported and then re-exported as the first app's own.

MacroFactor handles this with the priority rule above. Lose It's doc does not say what happens if you also write from another app, and I could not test it for this post.

Carbon Diet Coach took the blunt route: "We do not import nutrition from other apps." It reads back only its own entries, to avoid writing them twice.

Which fields does MealSaid write and what does it not do?

MealSaid writes five nutrition fields (dietary energy, protein, carbohydrates, total fat, fiber) and real workouts with active calories. It reads workouts, steps, weight, heart rate and sleep for context.

What it does not do:

• It does not read food from Health.
• It does not write water, caffeine, sugar, sodium or micronutrients.
• The Health data it reads never leaves your phone.

If you want 38 nutrients in Health, MyNetDiary or Cronometer will give you more than we do. If you want an app that imports what other trackers logged, that is MacroFactor. MealSaid's bet is narrower: get the meal into Health in one sentence, by voice or by Siri, with the app closed, and let Health be the ledger.

It also requires iOS 26, is English only, has no barcode scanner, and there is no Android version, so Health Connect users are out of luck.

How I checked this

Checked 20 October 2026. For each app I read the current Apple Health page in its own help center or, where none exists, the developer's own App Store description, and recorded the write and read lists and the page's "updated" date where shown. MyFitnessPal, Lose It and Cronometer block scripted fetches of their help centers (they return 403 to anything that is not a browser), so those three were read through the help center's public API and confirmed in Safari.

Cal AI's FAQ and listing were read the same day and found silent on nutrition writes. MealSaid's row is from our own HealthKit permission list. I did not install every app and inspect Settings > Health > Data Access for this post; that is a follow-up, with screenshots, and if it contradicts a doc I will say so here. Corrections to ohmbrock42@gmail.com.