Evidence over opinion.
Every food verified against USDA data. Every recommendation grounded in peer-reviewed research. Every macro target recalculated from your real progress. This is how Calibr8 stays accurate.
Every food, cross-referenced with USDA FoodData Central
Our database contains 4,700+ foods with 14 nutrient fields per food — macronutrients, minerals, and vitamins. Every value is cross-referenced with USDA FoodData Central, the gold standard in food composition analysis.
Automated 6-level validation scripts check every entry for calorie anomalies, missing values, and macronutrient inconsistencies. Each food passes manual review before it goes live. When USDA updates, we update.
1,316 peer-reviewed papers, curated by hand
Calibr8 ingests 1,316 peer-reviewed studies through an automated pipeline pulling from PubMed, Semantic Scholar, and OpenAlex. AI analysis scores quality and relevance, but never publishes on its own.
Scimago journal rankings (Q1 > Q2 > Q3 > Q4) sharpen the quality score — papers from top-tier journals carry more weight. The final selection of what lands on Calibr8 is always reviewed manually by a qualified nutritionist.
Your macros recalculate every week from your real data
Initial macros come from established sports science — Mifflin-St Jeor for basal metabolic rate, combined with individual activity and goal multipliers. Loss and gain rates are bounded by evidence-based floors and ceilings.
Every week the system recalculates your targets from real progress — weight changes, plan adherence, check-in feedback. Tissue-specific TDEE correction (Hall 2008) keeps long-term cuts and builds accurate. Static plans drift. Calibr8 doesn't.
Reviewed by a practicing nutrition coach
Hristiyan Marinov
Nutrition coach & Calibr8 founder
Science-trained nutritionist focused on evidence-based practice. Created Calibr8 to make accurate, adaptive nutrition coaching accessible to everyone — not just the ones who can afford a full-time coach.
LinkedInQuestions about our methodology
- How is food data verified?
- Every food in the database is cross-referenced with USDA FoodData Central and undergoes manual verification across 14 nutrient fields. Automated 6-level validation scripts catch anomalies before a food goes live.
- Where do research-backed recommendations come from?
- Recommendations are grounded in 1,316 peer-reviewed studies curated from PubMed, Semantic Scholar, and OpenAlex. AI analysis scores papers, but final selection is always manual by a qualified nutritionist.
- How often is data updated?
- The food database is refreshed weekly. The research pipeline pulls new publications Monday, Wednesday, and Friday. Nutrition formulas are reviewed whenever a new meta-analysis shifts the evidence.
- How do you protect against bias?
- Final research selection is manual and documented with source citations. Journal quality is weighted via Scimago rankings (Q1 outranks Q4), and Calibr8 prefers meta-analyses and systematic reviews over single studies.
The research, turned into your weekly plan
Personalized macros from a 1,316-paper evidence base and a 4,700+-food USDA-verified database. Adapts every week from your real progress.
Start the 7-day free trial7-day trial · Cancel anytime
Everything You Need
From a 4,700+ food database to weekly AI coaching, Calibr8 turns research into a plan you can act on.
Food Database
4,700+ foods with full macro and micronutrient profiles, including Bulgarian and international staples.
Browse foodsPersonalized Coaching
A weekly check-in turns your data into a plan that adapts to your progress, not a generic template.
See how it worksScience Guides
Long-form guides with citations on protein, fat loss, training nutrition, and supplements.
Read the guides