Cibo complete guide (2026): every screen, tool and feature of the quiet calorie tracker

Full disclosure, first paragraph, as always: Cibo is built by the same small Warsaw studio as HomeProof — the app whose blog you're reading — and Tunari VPN. The legal entity, TUNARIVPN Sp. z o.o., is printed in cibo.health's own footer with its Polish registry numbers. So this is not an independent review and we won't pretend otherwise. What we can offer instead is something an outside reviewer can't: a complete, accurate map of the product, its reasoning, and its honest limits. Every factual claim below comes from the public site itself, and where Cibo is simply average — AI accuracy on messy plates, for instance — we'll say so, because the product itself says so.

TL;DR — Cibo in one block

What Cibo is, and the problem it was built against

Every product category eventually develops a house style, and in calorie tracking that style is loud. Databases with eleven entries for the same apple. Streak counters that turn one missed Tuesday dinner into a broken chain. Badges for behavior that is, at its core, just eating food. The About page at cibo.health/about describes the founding irritation precisely: a tracker that "asks you to scan a barcode for a clementine," a streak that "makes Tuesday at 11 PM feel like failure," and newer AI apps whose interfaces are "built for showing off the AI, not for you eating dinner."

Cibo's answer is a design position rather than a feature list: AI where it actually saves time, classic where it shouldn't. Photo logging is AI because typing "grilled chicken with rice and roasted vegetables" into a search box is the single most annoying moment of food tracking. Portion editing is classic — a stepper that moves 10 grams per tap with the math updating instantly — because precision belongs to the human holding the plate. And the emotional register is deliberately adult: the app's job is to show you your numbers and get out of the way.

Three structural choices follow from that position, and they explain almost everything else in this guide:

• No shame mechanics. There are no streak-guilt loops or badge ceremonies. The one streak-like concept in the product — the group streak in Together — is shared and optional, a thing a circle protects together rather than a whip for individuals.
• Honesty about AI limits. Instead of pretending the model is always right, Cibo surfaces its uncertainty (more on the confidence system below). This is rare in the category and worth understanding in detail.
• Privacy as a default, not a tier. The studio's other products are a privacy-first document vault (HomeProof) and a VPN. The food diary inherits the same stance: never sold, never handed to ad networks, per the About page.

The studio behind it is worth a sentence, since it's us: a small team in Warsaw operating as TUNARIVPN Sp. z o.o., shipping three products on one operating principle — "make the boring infrastructure quietly intelligent so the user can get on with their life," as the About page puts it. If you've read our document-privacy playbook, you've seen the same philosophy applied to receipts and warranties.

The design language of quiet: four category anti-patterns Cibo refuses

To see what "quietly intelligent" means in practice, it helps to name the loud defaults of the category — four mechanics so common they feel inevitable, each of which Cibo deliberately omits:

• The individual streak. The chain of consecutive logged days is the category's favorite retention lever, and its failure mode is well known to anyone who's used one: the day the chain breaks, motivation doesn't dip — it collapses, and the app often gets deleted with it. Behavioral researchers call the pattern the "what-the-hell effect": one lapse reframes the whole effort as failed. Cibo ships no individual streak at all; the only streak in the product is Together's shared one, which a circle protects collectively and which exists to create support, not guilt.
• Badge inflation. Awards for logging a vegetable, logging a week, logging at all. The economics are simple — badges are free to mint — and so is the result: they devalue instantly and turn an adult tool into a sticker chart. Cibo's About page names this directly ("designed for a 14-year-old at a fitness camp") and ships zero badges.
• Notification pressure. The guilt-ping at 9 PM — "Don't forget to log dinner!" — treats the user as a lapsing patient. Cibo's social layer explicitly inverts it: "a wave instead of another notification," with the nudging delegated to people who actually know you, at their discretion, not an engagement scheduler.
• Confident nonsense. The quietest anti-pattern: AI estimates presented with uniform authority regardless of quality, because admitting uncertainty hurts the demo. Cibo's confidence levels, confirmation gates and zero-calorie block are the structural refusal of this one — covered in depth below.

None of these omissions is an accident of youth; all four are stated design positions on the public site. Whether that calm register retains users as powerfully as gamification does is a fair open question — but it is exactly the trade an adult audience keeps saying it wants.

The site journey, page by page

The rest of this guide walks cibo.health the way a curious visitor would — top navigation left to right, then down into the tools and resources most visitors never find. If you only have five minutes, read the homepage section and the Free-plan breakdown; if you're evaluating Cibo seriously, the calculator and Together sections are where the substance lives.

Stop 1: The homepage — a five-step promise

The homepage headline is six words — "Calorie tracking, made easy" — followed by the actual pitch: the easiest way to track calories with a photo; AI estimates calories, macros and portions in seconds; no searching food databases, no weighing, no guilt. Three badges under the download buttons set the commercial tone immediately: a Free plan exists, no card is needed to start, and you can cancel anytime. A fourth line — "Your data, never sold" — is the privacy stance stated before you've scrolled once.

Scroll one screen and you hit the numbers panel, which is the fastest way to understand the product's shape:

Below that, the homepage lays out the loop as five explicit steps, and it's worth quoting the structure because it is the product: (1) Capture your meal — take or upload a photo; (2) Identify the meal — the AI recognizes the food; (3) Edit portions & ingredients — fine-tune or swap; (4) See calories & macros — the full nutrition readout; (5) Sync to Today & track insights — save to your day and watch weekly trends build. Note where the human sits in that loop: step 3 is yours. Cibo's AI proposes; you dispose. The worked example on the page shows the readout style — a plate at 358 kcal with protein, carbs and fat each broken out both in grams and in grams-per-100-kcal, a density view that quietly teaches you which foods are doing the protein work.

The homepage then introduces Together ("the only calorie tracker you do not use alone"), an Apple Watch section — the About page calls the Watch "first-class because the wrist is the right place for a glance at the ring" — and closes with the standard download pair: App Store for iPhone and Apple Watch, Google Play for Android.

Stop 2: The six ways to log — in depth

Most tracker reviews list logging methods as bullet points. They deserve more than that, because the method you use 80% of the time determines whether tracking survives past week two. Cibo ships six, and each has a distinct job:

1. Photo logging (the headline act)

Point the camera at the plate; about five seconds later you have calories, macros and a portion estimate. The AI identifies the dish and its components — so a mixed plate comes back as its parts, each editable. Two details matter more than the headline speed. First, the estimate arrives with a confidence level (see the honesty layer below) rather than as fake certainty. Second, the result is a starting point designed for correction: tap into any component, adjust in 10-gram steps, swap an ingredient the AI misread, and the totals recalculate live. Photo logging draws from the monthly AI-log pool on the Free plan.

2. Voice logging

Say what you ate — "two eggs, toast with butter, and an orange juice" — and the AI parses it into logged items. Voice shares the same AI pool as photo and text. Its natural habitat: meals you've already finished, food eaten on the move, and any moment when photographing your plate would be socially odd (client lunches come to mind).

3. Text logging

The same parser, typed. Write a meal in plain language and the AI structures it. Where classic trackers make you hunt a database entry for every component — the About page's "scan a barcode for a clementine" complaint — text logging accepts the sentence you'd say to a friend and does the decomposition itself. Also part of the shared AI pool.

4. Barcode scanning (unlimited, even on Free)

For packaged food, scanning the barcode is strictly better than any AI estimate: the label data is the ground truth. Cibo's barcode lookup runs on Open Food Facts, the open, community-maintained database with 3M+ products, and — this is the notable part — barcode scanning is unlimited on the Free plan. It doesn't touch the 20-log AI pool. Cibo's own comparison page pointedly contrasts this with competitors where barcode access is "plan-dependent." If your diet is heavy on packaged goods, you can run Cibo meaningfully without ever paying.

5. Manual entry

Also unlimited on Free. The classic path: enter the food and amounts yourself. It exists because sometimes you know exactly what you ate — your own standard breakfast, a recipe you've computed before — and no AI middleman is needed. Combined with unlimited barcode, it means the Free plan's cap applies only to the convenience layer, never to the ability to track at all.

6. One-tap favorites

The repeat-meals accelerator. Most people's diet is dramatically more repetitive than they think — the same breakfast five days a week, the same three lunches on rotation. Favorites collapse those to a single tap, which in practice means the AI pool gets spent on genuinely new plates, not on re-estimating Tuesday's oatmeal. If you're optimizing a Free account, this is the lever: photograph a meal once, correct it carefully, favorite it, and replay it forever at zero quota cost.

Stop 3: The honesty layer — what makes the AI trustworthy enough to use

Here is the question every AI food tracker has to answer and most dodge: what happens when the model is wrong? Because it will be. Cibo's support page answers with unusual bluntness: the AI is "good on simple plates, decent on mixed ones — like every AI in this category." No magic claimed. What distinguishes the product is the machinery wrapped around that fallibility — four mechanisms that together form what we'd call the honesty layer:

1. Confidence levels on every estimate. Each scan tells you how sure the model is. A clean plate of grilled salmon reads differently than a casserole, and the interface says so instead of presenting both with identical authority. You learn, quickly, when to trust a result as-is and when to spend ten seconds correcting it.
2. Low-confidence results wait for confirmation. When the model is unsure, the result doesn't auto-save into your day. It parks, flagged, until you approve it. The default direction of error is "ask the human," not "pollute the diary."
3. A zero-calorie guess cannot be saved at all. A hard floor against the most corrosive failure mode — silent nonsense entering your data. If the model produces an obviously broken estimate, the product refuses to record it.
4. Cancelled scans refund your quota. If you abandon a scan, it goes back into your monthly pool. The pricing never charges you for the AI's misses — which also removes the perverse incentive to accept a bad estimate because "the scan is spent anyway."

Why dwell on this? Because data quality compounds. A food diary is only as useful as its worst entries: a week of trends built on a few silently absurd estimates is worse than no trends at all, since it teaches wrong lessons with a straight face. The honesty layer is Cibo's structural answer — uncertainty is surfaced, nonsense is blocked, and the human stays the editor of record. It's the same instinct that, in HomeProof, keeps OCR output as editable suggestions rather than silent writes to your vault.

Interlude: why the quiet design is also the evidence-based design

It would be easy to read Cibo's calm register as mere taste. The behavior-change literature suggests it's closer to engineering. Three findings, each mapped to a product decision:

• Friction kills habits faster than motivation builds them. The strongest single predictor of whether self-monitoring continues is how much effort each instance costs. Every step removed from the logging loop compounds across the hundreds of times you'll repeat it — which is the entire case for a five-second photo path, a one-tap favorites bench, and six methods so the lowest-friction one is always available for the situation you're actually in.
• Lapses are inevitable; what matters is their interpretation. Rigid all-or-nothing framing turns a skipped day into evidence of failure — the what-the-hell effect we mentioned above — while flexible framing treats it as noise in a long signal. A tracker with no individual streak has structurally removed the artifact that makes lapses feel catastrophic. You missed Tuesday; Wednesday's ring neither knows nor cares.
• Social support outperforms social comparison. Accountability to people you know personally is among the more durable adherence mechanisms; leaderboards of strangers are among the more fragile ones. Together is built on the first pattern — five known people, waves, a shared streak — and deliberately omits the second entirely.

None of this guarantees any individual outcome, and Cibo's materials are notably careful not to promise one. But if you're choosing between tracker designs, it's worth knowing that "quiet" here isn't an aesthetic refusal of engagement mechanics — it's a bet on the subset of mechanics with the better evidence behind them.

Stop 4: Together — the social layer that respects adults

At cibo.health/together lives what Cibo calls its signature feature: Together, "the calorie tracker for couples, families and friends." The premise, in the site's words: calorie tracking was designed for one person, but healthy habits rarely work that way. The implementation is a private circle of up to five people — and the design detail that elevates it is granular, per-person visibility control.

The Together page demonstrates with a sample circle: Alex (the host) shares summary, meals and photos with the circle; one member sees logged days, meals with calories, and streaks; another sees only logged days and streaks; a third sees just the summary. The point: you decide what to share and who sees it — person by person, not once for the whole circle. Your partner might see every meal with calories; your friend-of-a-friend in the same circle might see only that you logged today. Health data is intimate, relationships are asymmetric, and Together is one of the few social-fitness designs we know of that models that asymmetry instead of flattening it.

The interaction model follows the same quiet register as the rest of the app: "a wave instead of another notification." High-fives, kind notes, reactions to a win — lightweight gestures rather than algorithmic nudging. And the one streak in the product lives here: a group streak the circle protects together ("some streaks are easier to protect together," as the page puts it). Accountability as a shared project, not individual guilt — on a rough day, the circle seeing you show up is the mechanism, and the mechanism is other people, not a push notification.

Together ships on the Free plan. For couples running a household — the same people who keep a shared vault of receipts and warranties — it's the obvious entry point: one circle, two people, each controlling exactly what the other sees.

Stop 5: Apple Watch — the glance, done properly

The homepage gives the Watch its own section, and the About page states the reasoning in one line: Apple Watch is "first-class because the wrist is the right place for a glance at the ring." The division of labor is deliberate: the phone is where logging happens (camera, keyboard, barcode), and the wrist is where the day's state lives — your Today ring and remaining budget, available in the two seconds between raising your arm and deciding about the office pastry. Cibo ships on the App Store as an iPhone and Apple Watch app, and the Free plan's Apple Health sync means its data participates in the broader Health ecosystem you may already use.

Stop 6: The free tools suite — four calculators with their math on the table

Here is the part of cibo.health most visitors never find, and the part we'd recommend even to people who never install the app: four browser-based calculators under /tools/ — TDEE, macros, calorie deficit, and protein. No account, no email gate, and — unusually for this genre — the formulas are published right on the page, so you can check the arithmetic yourself. They form a sequence: TDEE tells you how much you burn, the deficit calculator turns that into a goal-shaped calorie target, the macro calculator splits the target into three daily gram numbers, and the protein calculator sanity-checks the number that matters most. Let's take them in order, because understanding the math is genuinely useful whether or not you ever use Cibo.

The TDEE calculator — how many calories you actually burn

TDEE — total daily energy expenditure — is the number every eating goal hangs from: the calories your body burns in a day of resting, moving and training. The calculator at cibo.health/tools/tdee-calculator computes it in two published steps.

Step one: BMR via Mifflin-St Jeor. Basal metabolic rate is what you'd burn lying still for 24 hours. Cibo uses the Mifflin-St Jeor equation — as the page notes, the formula most dietitians reach for because it outperformed the older Harris-Benedict equation in validation studies:

• Men: 10 × weight(kg) + 6.25 × height(cm) − 5 × age + 5
• Women: 10 × weight(kg) + 6.25 × height(cm) − 5 × age − 161

Step two: multiply by an activity factor — the standard clinical scale from ×1.2 (sedentary) through ×1.375 (lightly active), ×1.55 (moderately active), ×1.725 (very active), to ×1.9 (extremely active). The page's worked example: a BMR of 1,430 cal/day at "lightly active" lands on a TDEE of about 1,965 cal/day — eat around that and weight holds steady. It then derives three targets: a moderate cut (in the example, 1,465 cal/day, with the arithmetic shown that a ~500-calorie daily deficit corresponds to roughly 0.45 kg a week as an estimate, individual results vary), maintenance, and a lean-gain figure (~2,265 in the example).

One honest caveat the genre usually omits and we'll state plainly: every TDEE formula is an estimate with real error bars — activity multipliers are self-reported, and metabolisms differ. The practical use is as a starting number you then calibrate against 2–4 weeks of logged reality. Which is, of course, the argument for pairing a calculator with a tracker.

The macro calculator — three numbers instead of thirty

The macro calculator's framing is the best one-line explanation of macro tracking we've seen: turn a calorie target into "the only three numbers you have to hit each day" — grams of protein, carbs, and fat. It either derives your calories from the same Mifflin-St Jeor + activity pipeline or accepts a number you already know ("I know my calories"), then converts a percentage split into grams using the published energy values: protein 4 cal/g, carbohydrate 4 cal/g, fat 9 cal/g.

Four presets cover the standard philosophies — Balanced, High-protein (the page notes it's "favoured when cutting calories while trying to hold on to muscle"), Low-carb, and Endurance. The worked example on the page: a 1,465-calorie day at 40/30/30 yields 147 g protein, 110 g carbs, 49 g fat. And a thoughtful guardrail sits beneath the result: a protein sanity check that compares your computed protein against the sports-nutrition range for your body weight (in the example, 98–140 g/day, upper end favoured in a deficit) and tells you to switch to the high-protein preset if your split falls well under it. That's the difference between a calculator and a form: it argues with you when the output would be a bad idea.

The calorie-deficit calculator — the gap that drives change

The deficit calculator handles the subtraction between maintenance and target and makes the trade-offs visible: deeper deficits move faster but are harder to sustain and risk muscle; shallow ones are slower but livable. The widely used arithmetic — roughly 7,700 calories per kilogram of body fat — is why a moderate ~500/day deficit maps to about half a kilo a week on paper; bodies are not spreadsheets, which is exactly why the output should be treated as a hypothesis you verify with a few weeks of logging rather than a promise.

The protein calculator — the macro worth getting right

If you only track one macro, protein is the one: it protects muscle in a deficit, it satiates per calorie better than the alternatives, and it's the number most people quietly miss. The protein calculator produces a daily gram target from your body weight, goal and training, in line with the sports-nutrition ranges the macro tool references. Pair its output with the grams-per-100-kcal density view in the app's food readouts and you get a practical skill: seeing at a glance which foods actually carry protein and which just advertise it.

The formulas at work: three worked examples

Formulas are abstract until you run them. Here are three fully worked profiles through the exact pipeline Cibo's calculators publish — Mifflin-St Jeor, then the activity multiplier, then goal adjustment, then a 4/4/9 macro split. (All figures are estimates by construction; real-world calibration over a few logged weeks always outranks any formula.)

*Profile A illustrates why the calculators' guardrails matter: a flat −500 from a small sedentary TDEE lands uncomfortably low, and this is where a sane tool steers toward a gentler deficit, more daily movement to raise the TDEE side of the equation, or both. Profile C shows the protein sanity check doing its job — at 58 kg the sports-nutrition range lands roughly between 1.6–2.2 g per kg, so the gram target is set from body weight first and the remaining calories split afterwards. The general lesson across all three: the split percentages are the adjustable part; the protein floor is not.

Stop 7: The calorie database — 70 everyday foods, USDA numbers

At cibo.health/calories sits a browsable reference: calories, protein, carbs and fat for 70 common foods, shown both per 100 g and per typical serving — the double display matters, because per-100g lets you compare foods fairly while per-serving tells you what the apple in your hand costs. The numbers are sourced from USDA data (the U.S. Department of Agriculture's FoodData Central — the reference database dietitians use), and the entries read like a pantry: an apple at 52 cal/100 g and 95 per fruit, avocado at 160/100 g, banana at 89, strawberries at 32, watermelon at 30. It's an SEO page, obviously — but it's an SEO page that happens to be a legitimately handy kitchen reference, and it demonstrates the data standard the app itself leans on.

Stop 8: The comparison pages — how Cibo positions itself

Two pages, Cibo vs MyFitnessPal and Cibo vs Cal AI, show the product's self-image sharply — and notably, they're restrained by the standards of vendor comparison pages. The MyFitnessPal page opens by granting the premise ("both help you understand what you eat") and then stakes three claims: Cibo keeps the free plan broad, shows when its photo estimate is uncertain, and adds private Together circles. The table lists Cibo's annual price at $39.99 against "check current store price" for the competitor, barcode as "unlimited on Free" versus "plan-dependent," and a 7-day trial on the annual plan — with a dated footnote ("competitor information and prices checked Sept 2026") and an instruction to verify current terms before purchasing. Against Cal AI — the photo-first competitor — the differentiation shifts to the honesty layer and the social design: confidence shown per scan, and circles instead of a solo experience. We'd summarize the positioning in one sentence: Cibo doesn't claim to out-database MyFitnessPal or out-flash Cal AI; it claims a broader free tier, visible uncertainty, and the only private small-circle social layer in the category.

Stop 9: The blog — thirteen articles, zero fluff categories

The blog currently holds 13 articles under the tagline "backed by science, written for humans," sorted into tags that map the product's concerns: Nutrition (6), Switching (3), and one each for Together, Numbers, Privacy and How-to. The nutrition pieces are workflow-shaped rather than listicle-shaped — "Find Food Calories Fast Using USDA Data in 7 Steps," "5 Steps to Calculate Accurate Macros … With a Worked Example," "Estimate Recipe Calories in 10 Minutes," "Macro-Friendly Restaurants: 5 Easy Orders and How to Log Them," a barcode-reliability piece — and the Switching tag hosts the migration content for people arriving from other trackers. If you want a feel for the product's editorial voice before installing anything, the blog is the fastest read.

Six of the thirteen deserve a one-line map, because they double as a correspondence course in practical tracking:

• "Find Food Calories Fast Using USDA Data in 7 Steps" — the workflow for getting reference-grade numbers on any whole food: weigh, match the right USDA entry, log. The skill that upgrades every manual entry you'll ever make.
• "5 Steps to Calculate Accurate Macros, USDA Data and a Worked Example" — the long-form companion to the macro calculator: find BMR/TDEE, set protein first, then fat, fill the rest with carbs.
• "Macros for Fat Loss: A Calculator Workflow With USDA Data" — the deficit-specific variant: protect protein, log precisely, set macros by the math, review trends after 2–4 weeks rather than daily.
• "Estimate Recipe Calories in 10 Minutes Using USDA Data" — the home-cooking answer: weigh ingredients in grams, match raw-vs-cooked entries correctly, sum the batch, divide by servings. Includes the formula.
• "Macro-Friendly Restaurants: 5 Easy Orders and How to Log Them" — eating out without abandoning the week: protein-first orders, sauces on the side, a buffer for hidden fats, five reusable chain orders.
• "Scan Food Barcodes: 6 Quick Fixes and Why Calories Are More Reliable" — when barcode scans misfire and why label calories still beat estimates — useful context for the feature you'll use most on the Free plan.

Stop 10: Support — the FAQ that answers the awkward questions

The support page leads with a promise — write to the support address and "a real person will reply within 24 hours" — and a visible account-deletion path, which is both an App Store requirement and a trust signal when it's surfaced rather than buried. The FAQ is where Cibo's plainest language lives; three answers are worth quoting in substance:

• "How accurate is the AI, honestly?" — the answer we quoted in the honesty layer: good on simple plates, decent on mixed ones, like every AI in this category, with the four safeguards (confidence levels, confirmation gates, zero-calorie block, quota refunds) listed as what the product does about it.
• "What's the difference between Free and a paid plan?" — the full Free inventory (next section) followed by the sentence most paywalled products can't write: "Pro removes the 20-log cap. That's the whole difference — nothing else is locked."
• "Can I cancel anytime?" — yes; the subscription is managed by the store you bought it from, cancellation is "two taps, any time," and access runs to the end of the paid period.

The full Cibo food reference: 68 everyday foods, USDA numbers

Because this is a complete guide, here is the complete reference — every food on cibo.health/calories as of October 2026, reorganized by category with our reading of what each table actually teaches. All values are USDA-derived, shown per 100 g and per common serving (P = protein, C = carbs, F = fat, grams per 100 g). Treat it as the kitchen cheat-sheet it is.

Fruits

Fifteen fruits, none above 160 cal/100 g — and that one outlier is the avocado, which eats like a fat source (14.7 g fat) rather than a fruit. The pattern worth internalizing: berries and melon sit at 30–57 cal/100 g, making them the highest-volume snacks in the table, while the serving column shows how bananas (105/serving) and mangos (99) quietly double a fruit salad's total.

Vegetables

The cheapest calories on the page: ten of thirteen entries sit under 45 cal/100 g, with cucumber (15), zucchini (17) and tomato (18) nearly free by volume. The two exceptions that surprise people on plates: potato (77) and sweet corn / sweet potato (86 each) — honorary starches that belong mentally in the grains row.

Meat & Poultry

The protein engine room. Turkey breast is the table's efficiency champion — 30.1 g protein at 135 cal/100 g with 0.7 g fat — with chicken breast right behind (31 g at 165). The spread to watch is cut-to-cut: thigh carries three times the fat of breast, and bacon (541 cal, 42 g fat per 100 g) is a flavoring, not a protein strategy.

Seafood

Shrimp and canned tuna deliver 24–25.5 g protein at about 99–116 cal/100 g — the leanest protein per calorie anywhere in the database. Salmon and sardines cost more calories (206–208) but the difference is largely the fats you're often eating fish for in the first place.

Dairy & Eggs

The most bimodal category: egg whites and cottage cheese are lean-protein staples (the latter at 11.1 g protein per 84 cal), while butter (717 cal/100 g) tops the entire database. The serving column rescues cheese — 113 cal per cheddar serving is manageable; it's the per-100g figure (403) that explains how cheese boards escalate.

Grains & Legumes

Two stories in one table. Legumes — lentils, black beans, chickpeas — pair carbs with real protein (about 9 g/100 g), which is why they anchor plant-forward macro plans. Breads concentrate: 266 cal/100 g for white bread versus oatmeal's 71 — and the serving sizes (67 vs 166 cal) give you the honest per-slice, per-bowl math.

Nuts & Seeds

Every entry sits between 486 and 654 cal/100 g — the densest category by far, and the classic silent budget-breaker. The serving column is the discipline: an actual almond serving is 162 cal, a peanut-butter serving 188. Weighed portions matter more here than anywhere else on the page; the 10-gram stepper earns its keep.

Pricing: the Free plan, Pro, and the one-cap model

Cibo's pricing is unusually easy to describe, which is itself the point. Per the support page, the Free plan includes:

• 20 AI logs a month — photo, voice and text draw from one shared pool
• Unlimited barcode scanning and unlimited manual entry
• The Today ring and macro tracking — the core daily readout
• Together circles — the full social layer
• Apple Health / Health Connect sync

Pro removes the 20-log cap — and that's the entire difference. No features held hostage, no "advanced insights" tier, no locked export. The comparison page lists the annual plan at $39.99/year with a 7-day trial on annual; billing and cancellation live in the App Store or Google Play rather than a custom billing system, so cancelling is two taps and access persists through the paid period. (As always, verify current pricing on cibo.health or in the stores — terms change.)

Two honest observations about this model. First, the single-cap design means the Free plan is a genuine product, not a demo: with unlimited barcode + manual entry and favorites costing nothing, a disciplined user with a repetitive diet can track indefinitely at zero cost, spending AI logs only on novel plates. Second — the flip side — if photo logging is the whole reason you came, 20 logs is well under one per day, and the subscription is effectively the price of the convenience that attracted you. That is clearly the intended funnel, and it's a fair one because the uncapped alternatives genuinely work; but you should walk in understanding which user you'll be.

Privacy and data: the stance, stated and checkable

A food diary is more revealing than it looks. Logged meals encode your schedule, your health concerns, your discipline lapses, sometimes your religion, often your budget. Which is why the ad-funded pattern in this category — track for free, monetize the behavioral exhaust — deserves more scrutiny than it gets. Cibo's position is printed on its About page in one sentence: "your food diary stays yours: we never sell your data, and what you eat is never handed to an ad network."

Three structural facts support the claim beyond the sentence. First, the business model is aligned: revenue is the subscription, not advertising — the incentive to broker your dinner data simply isn't in the P&L. Second, the operator is a named EU company: TUNARIVPN Sp. z o.o., Warsaw, with NIP and KRS registry numbers published in the site footer — checkable in the Polish registry, subject to GDPR, reachable at a published address. That's a different accountability posture than an anonymous app-factory LLC. Third, account deletion is surfaced on the support page rather than hidden behind an email exchange. As we said about our own VPN in the VPN comparison: don't take any vendor's marketing line on faith, ours included — the policy itself lives at cibo.health/legal/privacy, and reading it takes five minutes.

Platforms: where Cibo runs

Worth stating because the category often fails at it: Cibo is genuinely cross-platform — a mixed iPhone/Android household (or Together circle) isn't a problem, and Health sync is native on both sides: Apple Health on iOS, Health Connect on Android.

Who Cibo fits — and who it doesn't

No tool is for everyone, and this guide has no credibility if it pretends otherwise. Our honest segmentation:

Cibo is a strong fit if…

• You've quit a tracker before because it felt like homework. The five-second photo loop plus favorites attacks exactly the friction that kills tracking in week two.
• You find streak mechanics counterproductive. If a broken chain has ever made you abandon an app entirely — the "what the hell" effect is real — a tracker with no individual streaks removes that failure mode.
• You're doing this with someone. Couples and families are Together's home turf, and the per-person visibility model respects that your partner and your coworker shouldn't see the same things.
• You care where your health data goes. Subscription-funded, EU-operated, no ad-network handoff — the privacy posture is the point, not a settings toggle.
• You live on packaged food or repeating meals. Unlimited barcode + favorites means the Free plan may simply be enough.
• You wear an Apple Watch. The wrist-glance design is built for you.

Look elsewhere if…

• You need the deepest possible food database with user-generated entries for every regional restaurant chain. MyFitnessPal's decade-old crowd-sourced database is still the widest net; Cibo's bet is that photo AI + Open Food Facts + USDA covers the practical cases, but "widest database" is not its claim.
• You want a large public community — forums, challenges, thousands of strangers. Together is deliberately the opposite: five people you actually know.
• You want coaching, meal plans, or a diet program built in. Cibo tracks; it doesn't currently coach. If you want an app that tells you what to eat, this isn't it.
• You photo-log heavily and won't pay. Twenty AI logs a month is the Free ceiling; past it, the honest options are barcode/manual/favorites or the subscription.

The household angle: food, receipts, and the same quiet philosophy

Why does a warranty-vault blog care about a calorie tracker beyond the shared parentage? Because the two apps sit on the same household workflow at different points, and the overlap is literal: the grocery receipt.

Run the thread: the receipt from Saturday's grocery run is, simultaneously, (a) a financial document — proof of spending, part of your household budget trail, the kind of thing worth keeping for the right number of years; and (b) a nutrition manifest — an itemized list of everything entering your kitchen, each line a barcode Cibo can scan at logging time. HomeProof captures the document the moment you unload the bags (on-device OCR, store, date, total extracted); Cibo captures the consumption as the week unfolds. Same quiet pattern both times: point the camera, let the AI do the typing, keep the human as editor, keep the data on your side of the fence.

The budget-conscious household gets a practical loop out of this pairing: HomeProof's receipts show what the food costs; Cibo's Insights show what the food does. Neither app needs the other — there's no forced bundle, no shared account, no cross-app tracking, which is exactly the point of the privacy posture both inherit — but a household that runs both ends up with the two ledgers that matter: what you own and spend, and what you eat. The third product in the family covers the pipe between you and the internet (the network layer), and the studio's thesis is complete: quiet, private infrastructure for the boring-but-vital parts of life.

Getting started: a realistic first week

If you decide to try Cibo, here is the first-week plan we'd actually recommend — designed around the Free plan's economics and the habit science, not around maximal feature usage:

1. Day 0 — set your numbers before the app. Spend ten minutes on the web calculators: TDEE first, then the macro split (run the protein sanity check). Write down three gram targets. Now the app has a job description.
2. Days 1–2 — photo-log everything, correct everything. Spend AI logs freely these two days. Your goals: calibrate your eye against the AI's portions (the 10-gram stepper is the teacher), and learn which of your plates read as high-confidence versus which need your edits.
3. Day 3 — build your favorites bench. Identify your repeaters — the breakfast you eat five days a week, the standard lunches — correct each one carefully once, favorite them. From here on, those meals cost one tap and zero quota.
4. Day 4 — switch packaged food to barcode. Label data beats AI estimates; it's unlimited; use it for everything with a wrapper.
5. Day 5 — invite your person. If you share a kitchen with someone, create a Together circle and set visibility deliberately — decide together what "accountability" means for you two before the defaults decide it for you.
6. Days 6–7 — read your first trends, adjust nothing. Two days of weekly-trend data is noise; resist the urge to re-plan. The discipline of week one is logging consistency, not optimization.
7. End of week — the honest checkpoint. Count your remaining AI logs. If you burned through most of 20 in a week and loved the photo flow, you're a subscription user — the 7-day trial on annual exists for exactly this moment. If favorites + barcode covered most meals, the Free plan may simply be your plan.

Beyond week one: a realistic month

Week one builds the logging reflex; the following three weeks are where the data starts paying rent. The same calibration logic Cibo's own blog recommends — review trends after 2–4 weeks, not after every dinner — turns into a simple monthly rhythm:

Week 2 — calibrate the formula against your reality

Your TDEE from the calculator was an estimate; your logged intake plus your scale's weekly average is the measurement. If two honest weeks of logging at your target moved your weekly average noticeably differently than the math predicted, believe the measurement: your real maintenance sits where your data says, and the fix is a small target adjustment (100–200 calories), not a new philosophy. This single habit — adjusting the number instead of doubting the method — is what separates people who track for a month from people who track usefully for a year.

Week 3 — make Together earn its slot

If you set up a circle in week one, week three is when the novelty fades and the design either works or doesn't. Two rituals worth trying before judging: agree with your circle on one shared check-in moment (Sunday evening, not daily), and use waves as the default reaction to a logged day — zero-effort acknowledgment beats commentary. If you skipped Together entirely, consider the two-person version with whoever shares your kitchen; the per-person visibility controls mean the experiment costs you exactly as much privacy as you choose.

Week 4 — audit the quota, decide the plan

Month's end is the honest accounting moment the 7-day trial couldn't give you. Pull up your month: how many AI logs did you actually spend, and on what? The common patterns: favorites carried me (your diet is repetitive — Free is genuinely your plan); I rationed photos and resented it (you're the subscription's target user, and $39.99/year is the price of removing that friction); I barely logged at all (no plan fixes that — revisit whether the week-one reflex ever formed, or whether this category of tool is for you at all). All three are legitimate outcomes, and a product confident enough to leave barcode and manual unlimited has, to its credit, made the third answer cost you nothing.

Glossary: the numbers Cibo throws around, defined

Frequently asked questions

What is Cibo?

An AI-assisted calorie tracker for iPhone, Apple Watch and Android from a small Warsaw studio: photo → calories, macros and portions in about five seconds, six logging methods in total, no streaks or shame mechanics, with a privacy-first data stance. The website is cibo.health.

How accurate is the AI photo logging, honestly?

Cibo's own words: good on simple plates, decent on mixed ones — like every AI in this category. The differentiator is the honesty layer: visible confidence per scan, confirmation gates on low confidence, a hard block on zero-calorie estimates, and quota refunds for cancelled scans.

What does the Free plan include, exactly?

20 AI logs a month (photo, voice, text — one shared pool), unlimited barcode scanning, unlimited manual entry, the Today ring and macro tracking, Together circles, and Apple Health / Health Connect sync. The paid plan removes the 20-log cap; per Cibo's support page, nothing else is locked.

How much does Pro cost?

The comparison page lists $39.99/year with a 7-day trial on the annual plan. Billing runs through the App Store or Google Play; cancelling takes two taps and access lasts through the paid period. Verify current terms on cibo.health or in the stores.

What is Cibo Together and how private is it?

Private circles of up to five people where visibility is set per person — one member might see your meals with calories, another only that you logged today. Waves and reactions replace notification pressure, and the only streak is a shared group streak. Included on the Free plan.

Does Cibo work without the AI features?

Yes, genuinely: barcode scanning (3M+ products via Open Food Facts) and manual entry are unlimited on Free, and one-tap favorites replay your corrected meals at no quota cost. A repetitive, package-heavy diet can be tracked indefinitely without touching the AI pool.

Who makes Cibo, and where does my data go?

TUNARIVPN Sp. z o.o., a registered Warsaw (EU) company — the same studio as HomeProof and Tunari VPN; the registry numbers are in the site footer. The stated stance: your data is never sold and never handed to an ad network. The specifics live in the privacy policy at cibo.health/legal/privacy — read it, as you should for any app holding health data.

Are the web calculators tied to the app?

No — the TDEE, macro, calorie-deficit and protein calculators run in the browser with no account, and they publish their formulas (Mifflin-St Jeor, standard activity multipliers, 4/4/9 macro energy values) on the page. They're useful even if you track in a notebook.

Cibo vs MyFitnessPal — the short version?

MyFitnessPal's strength is the widest crowd-sourced food database; Cibo's claims are a broader free tier (unlimited barcode on Free), visible AI uncertainty, and private Together circles. Cibo's own comparison table lists its annual price at $39.99 and — credit where due — tells you to verify the competitor's current terms rather than quoting a stale number as fact.

Is there a web app for tracking?

Tracking lives in the iPhone, Apple Watch and Android apps. The web presence is the marketing site, the calculators, the USDA food reference and the blog — there's currently no browser-based diary.

What is the difference between Open Food Facts and USDA data in Cibo?

Two different jobs. Open Food Facts is the open database of 3M+ packaged products behind barcode scanning — label data for things with wrappers. USDA FoodData Central is the reference source for whole foods — the chicken breast, the banana — and powers Cibo's /calories pages. Between them: packaged food gets label truth, whole food gets laboratory truth.

Can I use Cibo purely as a barcode scanner and manual tracker?

Yes — and at zero cost, since both are unlimited on the Free plan along with the Today ring, macros, Together and Health sync. The AI pool only matters if you want the camera-and-sentence conveniences.

I am switching from another tracker — where do I start?

The blog's "Switching" category exists for exactly this, and the practical path is the first-week plan above: set targets with the calculators, favorite your repeating meals in the first days, and let barcode handle your packaged staples. Your history stays in the old app; what transfers is your numbers, which the calculators regenerate in ten minutes.

Does the Free plan expire or require a card?

Per the homepage: there is no card to start, and the Free plan is a standing tier, not a trial. The 7-day trial is a separate thing — it applies to the annual paid plan, before the $39.99/year billing begins.

The verdict

We'll hold ourselves to the standard we set in our VPN comparison: a maker's guide you can trust is one that concedes real rows. So: Cibo will not out-database MyFitnessPal, it offers no coaching layer, its public community is deliberately five people large, and its Free plan caps the marquee AI feature at 20 logs a month. Those are the honest edges of the product.

Inside those edges, what the Warsaw team built is distinctive in three ways that survive scrutiny. The honesty layer — confidence levels, confirmation gates, the zero-calorie block, quota refunds — is a genuinely rare piece of AI product design in a category that prefers confident nonsense. Together's per-person visibility model treats adult relationships as asymmetric, which every other social-fitness design we've seen flattens. And the one-cap pricing — everything included, only the AI volume paid — produces a Free plan that is a real product and a subscription whose value proposition is simply "more of the thing you already like." Add the published-formula calculators, the USDA reference pages, an EU operator with its registry numbers in the footer, and a data stance of "never sold, never handed to ad networks," and you get what the About page promised: a calorie tracker for adults who'd rather just eat.

Start where we'd start: ten minutes with the TDEE calculator, then the app for a week on the plan above — and let your own logged reality, not any guide (including this one), tell you whether it stays.

Read next

• How long should you keep receipts? (the grocery-receipt angle)
• How to keep your financial documents private online
• Best VPN for everyday privacy: an honest comparison
• Best receipt scanner apps for iPhone (privacy compared)