Miriel
Download

Pediatric Nutrition AI

Pediatric nutrition AI, built for parents.

Miriel AI is the pediatric nutrition AI for parents of children from 6 months to 12 years — an iOS app that scans food labels with per-child safety scoring, generates AI meal plans tuned to each child's age, allergies, and growth stage, and supports sibling-aware planning across the whole family.

Support for babies starting solids (around 4–15 months) — first foods and early allergen introduction — is currently in beta; see the Starting Solids page. Miriel is for general guidance and educational purposes only, not medical advice; always talk to your pediatrician before introducing solids or new allergens.

Download free on the App Store How it works

What pediatric nutrition AI actually means

Most "nutrition AI" tools on the App Store are adult diet trackers with a smaller portion size for children — calorie counters scaled down, with allergen filters bolted on. Pediatric nutrition is a different problem. Children are not small adults; their biology is built around growth, not maintenance, and the safety constraints that matter for a toddler (choking hazards, iron deficiency, food protein reactions) are not the same as the constraints that matter for a 35-year-old tracking macros.

Pediatric nutrition AI is the category of AI tools designed for children specifically. The category's table-stakes capabilities are:

  • — Growth-percentile-aware energy and nutrient targets that follow WHO standards under age 2, CDC standards after age 2, and Korean standards for Korean families.
  • — Per-child allergen and condition safety evaluated as hard constraints, not preferences to be over-ridden by convenience.
  • — Age-appropriate choking-hazard rules for whole grapes, whole nuts, hot-dog rounds, and other shape-modified foods under age 4.
  • — Developmental-stage awareness — for example, no fat restriction under age 2 (which is essential for brain development), no calorie deficits for children at all.
  • — Picky-eating and acceptance modelling grounded in feeding-behaviour research (Birch, Satter), so plans expand a child's diet through repeated low-pressure exposure rather than coercion.

The clinical references behind every Miriel recommendation are listed openly on the research page. The methodology — how Miriel combines those references with AI models in a specific safety-first order — is described on the methodology page.

Why adult nutrition AI does not work for children

The fastest way to see the gap: open any adult-targeted nutrition app and add a 2-year-old as a profile. The defaults are wrong. The portion sizes are adult portions scaled to a smaller mass. The fat targets are adult guidelines (which would be harmful under age 2). The allergen filters are filters, not first-class data — if you uncheck "peanuts" the app still surfaces peanut-containing recipes and asks you to remove them manually. There is no choking-hazard logic. There is no growth-percentile context, because adult apps don't think in growth at all.

Miriel inverts this. The defaults start from pediatric standards — current WHO, CDC, and Korean growth references, US National Academies and Korean DRIs for nutrient targets, AAP Bright Futures for clinical milestones, the LEAP and EAT studies for allergen introduction timing. Adult assumptions are not allowed to leak into the child's recommendation.

What Miriel does, concretely

Food-label scanning with per-child safety

Point your iPhone camera at a packaged food. Miriel reads the ingredients and the nutrition panel, then returns a safety score specific to your child's profile — allergens flagged (including may-contain warnings), age-appropriate portion guidance, hidden added sugar quantified.

AI meal planning, child-specific

Tell Miriel what is in your kitchen and which child you are cooking for. Miriel generates a meal plan for a meal, day, or week, respecting your child's allergies, age, growth stage, dietary preferences, and what they have actually eaten before. For multi-child households, the plan computes per-child portions and modifications from a single shared recipe — one meal, sibling-aware.

AI nutrition coach

Chat with an AI nutrition coach that knows your child's profile. Premium voice mode is available on the Standard plan (with a daily allowance) and on Advanced (real-time, with a higher daily allowance). The coach is clear about its limits — it does not diagnose, it points you to a pediatrician when a clinical answer is what the question actually needs.

Growth tracking and reports

Track your child's growth against WHO and CDC standards (Korean standards for Korean families). Weekly and monthly reports show trends, percentile movement, and nutrient adequacy patterns.

Built for these families specifically

Pediatric nutrition AI — common questions

What is pediatric nutrition AI?

Pediatric nutrition AI is artificial intelligence built specifically for the nutrition needs of children — typically from around 6 months through age 12 — rather than adapted from adult nutrition tools. It handles things that adult AI cannot do safely without modification: age-appropriate portion sizes, choking-hazard rules by age, per-child allergen alerts, pediatric growth-percentile context (WHO under 2, CDC after 2), and developmental considerations like the under-2 fat-restriction reversal. Miriel AI is the pediatric nutrition AI for parents.

How is pediatric nutrition AI different from adult nutrition AI?

Adult nutrition AI is usually built around weight goals — losing or maintaining weight — with calorie deficits, fasting protocols, and macronutrient targets calibrated to adults. None of that applies to children, whose biology is built around growth, not maintenance. Pediatric nutrition AI inverts the defaults: growth-percentile-aware energy targets, age-tuned macro and micronutrient needs, allergens treated as hard constraints, choking-hazard rules that adult tools never check, and explicit refusal to generate calorie-deficit weight-loss plans for children.

Is pediatric nutrition AI safe?

Pediatric nutrition AI is as safe as the standards it is built on and the safety constraints it enforces. The right pediatric nutrition AI cites its references publicly (Miriel does this on /research), enforces allergen and choking-hazard rules as hard constraints that no other layer can override, and is explicit about what it does not do — diagnosis, treatment, replacing pediatric or dietetic professionals. Miriel publishes its methodology openly at /methodology so families and clinicians can verify the approach.

Can pediatric nutrition AI replace a pediatrician?

No — and any pediatric nutrition AI that claims it can should be treated with extreme caution. Miriel does not diagnose medical conditions, does not replace allergy testing, does not provide clinical advice, and does not generate weight-loss plans for children. It supports the day-to-day work of feeding a child within whatever care plan a pediatrician or registered dietitian has set.

Who is pediatric nutrition AI for?

Parents of children from 6 months to 12 years who want pediatric-specific guidance — not a downsized adult app. The use cases include: families navigating picky eating, families with food allergies or medical conditions (celiac, EoE, CMPA, FPIES), parents of toddlers in the iron-deficiency window, multi-child households where one meal needs to work across siblings of different ages, and families with cultural cuisine preferences (Korean, vegetarian, vegan, halal, kosher) that need real adaptation, not surface filtering.

How much does pediatric nutrition AI cost?

Miriel is free to download with a usable free tier — one child, one food scan and five AI coach chats per day, basic nutrition logging. Standard ($4.99/month or $49.99/year) removes the daily limits for one child and adds reports and growth tracking; Advanced ($9.99/month or $99.99/year) supports unlimited children on a single account with deeper analytics and full feature access. Full details on /pricing.