Noom, the psychology-based weight-loss app that built its reputation on cognitive-behavioral coaching rather than calorie math, has spent 2026 expanding its AI features to compete in a nutrition-tracking market increasingly defined by photo-based food logging and GLP-1 medication management, according to the company’s own announcements and industry coverage.
From human coaches to AI-assisted logging
Noom’s core model has always paired human coaches with behavioral-science content, but the company has layered in AI tools to make daily logging less tedious: users can now log meals via photo, text, or voice, with an AI system identifying foods and estimating nutritional content automatically. Noom describes its AI-enabled chatbot as a complement to human coaching, available around the clock for quick questions, while coaches remain the ones guiding longer-term behavior-change strategy.
Clinical evidence sets Noom apart
What distinguishes Noom from newer AI-native diet apps is its clinical evidence base: the company points to more than 40 peer-reviewed publications, including a randomized controlled trial in which 75 percent of participants maintained at least 5 percent body-weight loss at one year — a result well above what most self-directed dieting apps report. That evidence base has let Noom expand into Noom Med, its telehealth arm offering GLP-1 prescriptions alongside coaching, positioning the app as a hybrid between behavioral therapy and medical weight management.
Pricing and competitive pressure
Noom Weight is priced at roughly $17 a month on a 12-month plan, undercutting some AI-native competitors while still charging a premium over free calorie-counting apps. That pricing sits in a market being reshaped by rivals: MyFitnessPal acquired the AI photo-logging app Cal AI in a deal that closed in December 2025 and was announced in March 2026, folding a dedicated AI food-recognition tool into its own Premium tier. The acquisition underscores how central AI photo logging has become to the category — a feature Noom has had to match rather than pioneer.
Where AI food-logging still falls short
Independent testing of AI food-recognition tools across the category has found meaningful accuracy gaps: one benchmark found a leading photo-logging tool correctly identified only about 71 percent of food items in a 500-image test set, with accuracy dropping to roughly 58 to 61 percent for East Asian and South Asian cuisines. Those numbers illustrate a broader limitation across AI nutrition apps, Noom included: photo recognition trained predominantly on Western food images tends to underperform on other cuisines, meaning users still need to manually correct a meaningful share of AI-generated entries.
Noom’s approach also reflects the company’s origin story. Founded on cognitive-behavioral-therapy principles rather than pure calorie counting, Noom has always tried to differentiate itself from bare-bones tracking apps by emphasizing psychology-driven habit change over raw numbers. That positioning is partly why the company has been slower than Cal AI or Dexcom’s Stelo to lean fully into AI-first food recognition — it is layering AI onto an existing coaching relationship rather than building the app around AI logging from scratch, a distinction Noom has continued to emphasize as competitors race toward fully automated tracking.
The GLP-1 medication boom has reshaped Noom’s business in a way that goes beyond software. Noom Med’s ability to prescribe GLP-1 drugs alongside behavioral coaching puts it in more direct competition with telehealth-only prescribers, and the company has increasingly framed its AI logging tools as a way to help GLP-1 patients track appetite changes and nutritional gaps that can arise when medication drastically reduces food intake — a use case behavioral coaching apps without a medical arm cannot easily replicate.
What it means for dieters
The shift toward AI-assisted logging is lowering the friction of daily tracking, which behavioral research has long identified as the biggest predictor of whether a diet program sticks. But the accuracy gaps mean AI logging works best as a shortcut for rough tracking rather than a substitute for careful nutrition counting, particularly for non-Western diets. As GLP-1 medications reshape how many people approach weight loss, expect Noom, MyFitnessPal, and Dexcom’s Stelo — all racing to add AI nutrition features in 2026 — to keep competing on which combination of accuracy, coaching, and medical integration wins users’ long-term loyalty. For dieters choosing between apps, the practical takeaway is that no current AI food-logging tool — Noom included — should be treated as fully hands-off; spot-checking AI-generated entries against a food label or database remains worthwhile, especially for anyone eating a cuisine underrepresented in these companies’ training data.
