Clinically-validated · Zurich, CH

One snap. The full carb count.

Snaq photographs your meal, estimates carbs, protein and fat in seconds, and pins the result to your glucose trace — so you can finally see which meals work and which spike.

Get Snaq free See the glucose link
4.5 ★
App Store
250k+
Users
40%
T1D
RCT
Validated
CURRENT GLUCOSE
118mg/dL
In range
10:00 11:30 13:00 14:30 16:00
Chicken bowl, brown rice
Logged 12:45 · Photo
47g carb
28g pro
12g fat
Photo carb countingDexcom integrationFreeStyle LibreTime-in-RangeAI NutritionistApple Health syncLiDAR scanningHCP reportsSaved meals Photo carb countingDexcom integrationFreeStyle LibreTime-in-RangeAI NutritionistApple Health syncLiDAR scanningHCP reportsSaved meals
How it works

Three steps from plate to glucose insight.

The legacy carb-counting workflow was math and guesswork. Snaq is a photo, an AI estimate, and a glucose trace that closes the loop.

1

Snap the plate

Open the app, frame the meal, take a picture. On iPhone Pro the LiDAR sensor adds depth data for better portion estimation. Barcode scan works for packaged items.

≈ 2 seconds
2

AI does the math

The food-recognition model identifies ingredients, estimates quantities, returns carbs, protein, fat and total calories. Adjust if the estimate is off, then save.

≈ 4 seconds
3

See the glucose link

Your CGM stream pins the meal onto the trace. After 2 hours you have a clear answer: did this meal stay in range, or did it spike? Trends compound over weeks.

2 hr post-meal
Killer feature

Every meal, pinned to a glucose curve.

The reason a photo-carb counter belongs in your diabetes stack: it closes the loop between what you ate and what your body did with it.

Snaq pulls your CGM trace from Dexcom, FreeStyle Libre, Abbott, Medtronic or whichever sensor you use, then anchors each logged meal to the moment you ate it. Two hours later, the area-under-curve is yours: was the spike steep, did you stay in your personal target range, did the macro mix help or hurt? Over a few weeks, the patterns become readable — and that is where the Time-in-Range score actually moves.

Honest note: Snaq is a wellness app, not a regulated medical device. The carb estimate informs mealtime decisions; insulin dosing should follow your clinician's plan and your own judgment, especially on meals the AI is uncertain about.
8%Low
72%In range
20%High
< 70 mg/dL 70–180 mg/dL > 180 mg/dL
7-day Time-in-Range 72%+6%
Meals logged 38this week
Avg post-meal peak 142mg/dL
Who it's for

Built for diabetes-aware eaters.

40% of Snaq users live with T1D, 40% with T2D, the rest with pre-diabetes or as caregivers. The app is shaped around those four use cases.

T1D

The Type 1 carb counter

You bolus around meals, every gram matters. Snaq gives a fast estimate plus the post-meal trace to verify whether your dose covered it.

T2D

The Type 2 patterner

You may not be on insulin, but glucose patterns matter for medication choices and lifestyle calls. Snaq surfaces which meals reliably spike.

PRE

The pre-diabetes early-mover

An HbA1c in the 5.7–6.4 range and a CGM trial. Use Snaq to learn your individual glycemic response before any of it hardens.

CARE

The caregiver / parent

You log meals for a child, partner or parent. Shared reports go straight to PDF or spreadsheet for the next diabetes appointment.

Features

A diabetes-shaped nutrition stack.

Not a generic calorie counter with health features bolted on — every tool was designed against the specific question "how did this meal move my glucose?"

Photo-to-carbs in seconds

Snap any plated meal; the AI returns carbs, protein, fat and calories. LiDAR adds depth on iPhone Pro. Editable estimate before save. Learns your repeats.

S

Barcode scanner

Packaged-food lookup with verified nutrition labels. The most precise reading for processed items.

AI Nutritionist coach

Chat-style suggestions based on your recent log and glucose response. Swaps, additions, portion adjustments.

CGM & sensor integration

Pulls glucose, insulin and activity from the major sensors and apps.

Dexcom FreeStyle Libre Abbott Medtronic mySugr Tidepool Apple Health

HCP-ready reports

Export PDF or spreadsheet of meals, glucose and trends for your clinician.

Saved meals & one-tap recall

Re-log a frequent meal in a single tap — the same dish always gets the same carb estimate, and the AI gets smarter on your usual repeats.

M
Compare

Snaq vs the other options.

Honest matrix — including the cells where Snaq loses. If mySugr's gamification or Cronometer's micronutrient depth fits you better, that's the right call.

Feature Snaq mySugr Tidepool MyFitnessPal
Photo-to-carb AI ✓ Native ✗ ✗ ◐ Premium
CGM integration ✓ 7+ sensors ✓ Multiple ✓ Aggregator ✗
Glucose-meal correlation ✓ Auto-pinned ◐ Manual ◐ Clinician view ✗
Gamified BG logging ◐ Charts only ✓ Monster mascot ✗ ✗
Food database size ◐ AI-driven ◐ Mid-size ✗ ✓ Largest catalogue
Micronutrient depth ◐ Macros + key ◐ Basic ✗ ◐ Basic
Clinical validation (RCTs) ✓ Published ◐ Industry ✓ Clinician-grade ✗
Free tier limits ◐ 5 meals/week ✓ Generous ✓ Free ✓ Generous
Annual price (web) ✓ $60 ◐ $35 ✓ Free ◐ $80

Pricing and feature data compiled from public product pages in May–June 2026. Verify on the developer's site before subscribing.

From real users

What three different diabetic users say.

From public App Store and Google Play reviews. Including one mixed-rating because every CGM-adjacent tool has an honest trade-off.

★★★★★
T1 for 48 years. Effortlessly photo, barcode scan, search huge food database. Photo feature is quite accurate. Smooth integration with the Libre 3 CGM. Extensive reports to PDF or spreadsheet.
B
Brian H.
Google Play · Type 1 · Libre 3
★★★★★
Pre-diabetes, got a CGM. Pairing was fast, AI coach is genuinely useful. But photo recognition needs my sanity check half the time — soups and wraps are guesses. Worth it overall.
P
Poonam A.
Google Play · Pre-diabetes
★★★★★
T2 since 2019. Snaq makes the connection between meals and glucose visible in a way mySugr never did for me. AI Nutritionist suggestions actually shifted my breakfast routine.
M
Marcus L.
App Store · Type 2 · Dexcom G7
Story

From a Zurich lab, with peer review.

SNAQ GmbH was founded in 2019 in Zurich by Aurelien Briner and co-founders, originally to crack one specific problem: carb counting for Type 1 diabetes was tedious, error-prone and stuck on manual databases. The bet was that food-recognition computer vision had matured enough to make a photo workflow viable for clinical-grade decisions.

What separates Snaq from the wider AI-tracker crowd is the clinical evidence layer. The team has published peer-reviewed accuracy studies on portion estimation and run randomized controlled trials on Time-in-Range outcomes — an unusually high evidence bar for a consumer health app. The advisory board includes diabetes clinicians and engineers. After the US launch in 2022 the app reached around 250,000 users with 10–20% month-over-month growth.

Honest disclosure: the photo AI is strong but not perfect — community reviews flag that soups, wraps and heavily stacked food can confuse it, and you sometimes can't get the model to recall a meal from a similar-looking photo (only from a saved meal). The free tier is hard-capped at 5 meals per week, which is tight if you want full-day logging without subscribing. Google Health Connect sync has had reported flakiness in 2025–2026.

The Zurich HQ and Swiss/EU base mean GDPR-grade data handling and a privacy portal designed for export and deletion. Snaq is a wellness tool, not a regulated medical device — the carb estimates feed your judgment and your clinician's plan, not insulin dosing automation.

Quick facts

  • Founded2019, Zurich
  • Legal nameSNAQ GmbH
  • CEO & Co-founderAurelien Briner
  • Users~250k
  • User mix40% T1D · 40% T2D
  • App Store rating4.5 ★
  • CGM partnersDexcom · Libre · more
  • PlatformsiOS · Android
  • EvidencePeer-reviewed · RCT
  • Data baseSwitzerland / GDPR
FAQ

The nine questions that come up.

For anything not covered here, check the privacy and product pages at snaq.ai or open a support request from inside the app.

What is Snaq and how does the photo carb counter work?
Snaq is an AI diabetes food tracker. You take a photo of your meal and the app's food-recognition model identifies the ingredients on the plate, estimates portion sizes, and returns carbs, protein, fat and calories in a few seconds. On iPhone Pro models with LiDAR, the app uses depth data to improve portion estimation. The result is editable — you can adjust ingredients or quantities before saving. Snaq is built for people managing Type 1, Type 2 and pre-diabetes, and connects the food log to glucose data from a CGM.
Which CGMs and diabetes devices does Snaq integrate with?
Snaq integrates with Dexcom (G6, G7, ONE+), FreeStyle Libre (2, 3, 14-day), Abbott LibreView, Medtronic CareLink, mySugr, Sugarmate, Tidepool, Ascensia Diabetes Care and Apple Health. On iOS, Snaq can both read glucose, insulin and activity data, and write carbohydrate, fat and protein values to Apple Health. Connection is one-time OAuth and re-syncs automatically after major OS updates, though Health Connect on Android has had reported sync issues in recent months.
How accurate is Snaq's photo carb counting?
Snaq has published peer-reviewed accuracy studies and randomized controlled trials, which is unusually transparent for this app category. In practice the AI works best on well-lit, on-a-plate meals — soups, mixed bowls and stacked food are harder, and the model sometimes mis-identifies similar-looking foods. Treat the result as a strong starting estimate that still benefits from a sanity check, particularly if you are using the carb number to dose insulin. The accuracy improves over time as you save repeat meals and the app learns your patterns.
What is the AI Nutritionist coach?
The AI Nutritionist is a chat-style coach inside Snaq that reviews your recent meal log and glucose response, then suggests adjustments — for example, swapping a high-glycemic side, splitting a large portion, or adding a small protein or fiber buffer before a known spike-trigger meal. It is meant as a personal-tutor layer over the raw tracking data, not as a substitute for medical advice from your doctor or diabetes educator.
How much does Snaq cost?
Snaq has a free tier limited to 5 meals per week, which is fine for sporadic logging but tight for full-day tracking. Premium unlocks unlimited meal logging, CGM integration, AI Nutritionist and personalized insights. Web pricing is the cheapest: $2.66/month on a 3-year plan ($96 total), $5/month annually, or $9/month on a 3-month plan. App Store and Google Play subscriptions may cost more because of store fees. Web purchases have a 30-day money-back guarantee processed through Stripe. App store purchases follow Apple's and Google's refund policies.
Is Snaq a medical device? Can I use it to dose insulin?
Snaq is a health and wellness app, not a regulated medical device or clinical decision-support tool. The carb estimates can inform mealtime decisions, but insulin dosing should follow your healthcare provider's plan and your own judgment, especially when the AI's confidence on a particular meal is uncertain. Snaq's exportable PDF and spreadsheet reports are designed for review with your doctor, diabetes educator or dietitian — not as a replacement for them.
How does Snaq compare to mySugr, Tidepool or general calorie counters?
mySugr (now owned by Roche) is strong on gamified blood-glucose logging and insulin tracking, weak on food estimation — you typically type meals manually. Tidepool excels at multi-device data aggregation for clinicians but is not built around meal logging. General trackers like MyFitnessPal and Cronometer have huge food databases but no CGM integration and no glucose correlation. Snaq's lane is the specific intersection: photo-to-carb in seconds, glucose correlation per meal, and clinical evidence behind the estimates. If you do not have a CGM, a general tracker may suit you better.
Does the photo recognition work on home cooking and restaurant meals?
Yes for plated, visible meals. The model handles home-cooked plates, restaurant dishes shot from above, mixed bowls and packaged items reasonably well. It struggles when food is heavily stacked, when items overlap colors, when lighting is poor, or when you photograph a wrap or burrito where the fillings are hidden. For drinks, sauces and complex layered dishes a quick edit usually closes the gap. The barcode scanner handles packaged products and gives the most precise read for those.
What about my data — is glucose and meal data kept private?
SNAQ GmbH is a Swiss company subject to Swiss data protection law and GDPR for EU users. The privacy policy at snaq.ai covers data categories (account, body metrics, food images, CGM streams, insulin doses), retention windows and third-party processors. Photos may be retained and used to improve the food-recognition model with opt-out controls in settings. PDF and spreadsheet exports are generated on demand and are not auto-shared with any third party. You can export and delete your account through the privacy portal.
Get the app

Start free. Find out which meals actually spike you.

Photo carb counting, CGM correlation and the first taste of the AI Nutritionist are unlocked on the free tier. Premium opens unlimited meals.

Download Snaq Try Premium 7-day free

iOS 15+ · Android 9+ · 30-day money-back on web · Cancel anytime

Health disclaimer: Snaq is a wellness tracker, not a regulated medical device, not a clinical decision-support tool, and not a substitute for professional medical advice. Do not adjust insulin doses or other medications based solely on Snaq's carb estimates. Always consult your endocrinologist, diabetes educator or primary care clinician for treatment decisions.
S
Snaq — Diabetes Food Tracker
PHOTO · CGM · FREE
Get