Photos plus CGM, with the AI on top.
Undermyfork shows you photos next to glucose curves. SNAQ does that and tells you what to change next time, with the AI doing the work.
In short.
Undermyfork pairs meal photos with CGM data and calculates Time in Range per meal, with a small but loyal T1D-focused user base. SNAQ does the same photo plus CGM pairing and adds three things: AI-derived nutrient estimation from the photo (validated at 5.5 g mean absolute carb error), predictive glucose response before the meal, and a randomised controlled trial showing improved Time in Range.
At a glance.
- Undermyfork's core is meal photos plus automatic Time-in-Range calculation per meal. SNAQ adds AI-derived carb and macro estimation, glucose prediction, and proactive coaching.
- SNAQ photo carb counting is validated at 5.5 g mean absolute error. Undermyfork's photo recognition is tag-based, not real macro estimation.
- SNAQ has a 2025 RCT (+6.6 pp Time-in-Range). Undermyfork's evidence is a 2022 pilot study presented at the Diabetes Technology Meeting.
- Both serve T1D users; SNAQ also covers T2D, prediabetes and GLP-1.
- Undermyfork has its own provider portal; SNAQ Care is used by 500+ clinicians.
Feature by feature.
Structured comparison based on publicly available information, last reviewed May 2026.
| Feature | SNAQ | Undermyfork |
|---|---|---|
| Meal photo + CGM overlay | ✅ | ✅ Core feature |
| Sensor-optional use | ✅ Works fully without a CGM or BGM; supports intermittent sensor wear | ❌ Sensor required for the Time-in-Range value prop |
| AI carb and macro estimation from photo | ✅ Validated at 5.5 g mean absolute carb error | ⚠ Tag-based, manual |
| Glucose prediction before the meal | ✅ | ❌ |
| AI Nutritionist coaching | ✅ Proactive | ❌ |
| Peer-reviewed clinical evidence (product-specific) | ✅ RCT, eClinicalMedicine 2025 | ⚠ 2022 pilot, Diabetes Technology Meeting |
| Cohort coverage | T1D, T2D, prediabetes, GLP-1 | T1D primarily |
| Provider dashboard | ✅ SNAQ Care, 500+ clinicians | ✅ Undermyfork Care |
| Best for | T1D, T2D, prediabetes, GLP-1 users who want AI guidance | T1D users who want simple photo + CGM history review |
Sources: Herzig D. et al., JMIR mHealth and uHealth, 2020 (carb counting accuracy) Piazza et al., eClinicalMedicine 2025 (RCT NCT05671679, Time-in-Range) SNAQ research overview Undermyfork product pages
Why users switch.
AI carb counting, not photo tags
Undermyfork shows you the photo and lets you tag what you ate. SNAQ recognises the meal and estimates the carbs for you, with validated accuracy.
Forecast, not just hindsight
Time-in-Range per meal is retrospective. SNAQ also tells you what's likely to happen for this meal, before you eat it.
Wider cohort coverage
Undermyfork is narrowly T1D-focused. SNAQ supports T1D, T2D, prediabetes and GLP-1 cohorts in one product.
Honest take.
Choose SNAQ if…
- You want AI to estimate the carbs and macros, not to tag them yourself.
- You want predictive guidance before you eat.
- You may benefit from GLP-1 or T2D modules alongside T1D features.
Consider Undermyfork if…
- You only want a visual photo plus CGM history without AI features.
- You're a heavy MiaoMiao or BluCon user and prefer Undermyfork's narrow T1D focus.
Bottom line
Undermyfork shows. SNAQ shows and recommends. Both rely on photos plus CGM, only one closes the loop with AI.
Published randomised controlled trial evidence, measured against standard care in T1D patients on automated insulin delivery systems.
Source: eClinicalMedicine (The Lancet), 2025 · Piazza et al. · NCT05671679
Frequently asked.
What's the main difference between SNAQ and Undermyfork?
Is the clinical evidence for SNAQ independently verified?
Which CGMs and devices does SNAQ work with?
Does SNAQ support FreeStyle Libre through MiaoMiao or BluCon?
See the difference yourself.
No logbook can coach you. Download SNAQ free and find out what your data has been trying to tell you.
4.6 · 3,000+ App Store ratings