TrainersEye
MoPro
MoPro Codex is the exercise ontology, biomechanics library and API behind the TrainersEye apps. We license it to universities, research institutions and fitness and health apps — through the API or as a data licence.

Counted on 12 November 2026 from the live MoPro Codex API, the blueprint and comparison catalogues, the Q&A store and the figure media store. Variations are generated candidates and stay labelled that way until experts review them.
Our main goal is partnership: research and teaching licences, review collaborations and commercial API licences.
One exercise vocabulary for kinesiology, physiotherapy and sport-science teaching, 600 blueprints for coursework, the Open Library for your learning platform, and supervised projects where students review candidate variations.
→🔬Stable exercise identifiers for trials and datasets, an ontology to annotate video and motion data, consistent exercise figures for papers, and validation studies that move variations from candidate to reviewed.
→📱Search that understands gym shorthand, variation trees for regressions and progressions, exercise illustrations in three sizes, and an evidence and safety boundary on every response.
→Base URL api.mopro-codex.com, Bearer-key authentication and an OpenAPI 3.1 specification. The API is in private developer beta.
Profiles, parent–child variation trees, children and aliases for 7,000 base exercises.
/v3/exercise/profile · tree · children · aliasesTypo-tolerant search across canonical names, aliases and 58,490 search terms, with a confidence score.
/v3/exercise/search600 movement blueprints with a web view and PDF for each, in 40 movement families.
/reference/{slug}.html · .pdfFigure requirements and status per exercise, and illustrations for 19,236 exercises.
/v3/exercise/figuresMovement-science answers grounded in the MoPro knowledge base, in fast and in-depth modes.
/v3/chat/askModelled biomechanics emphasis — joint angle, torque, phase timing, stability demand and more.
/v3/exercise/graph-domains7,000 base exercises with canonical names, movement families, 14,239 aliases and 58,490 search terms.
28,000 generated variations are marked as ready for expert review. We never present them as validated.
Graph-domain outputs are modelled interpretations for educational decision-support — not medical diagnosis.
Learning modules, slide decks, 2,500 exercise comparisons, research readings and the 600 blueprints, on the web and inside the TrainersEye app. Titles marked CC BY 4.0 are free to share with attribution.
Universities, research groups and health companies: email us your organisation, your project and your timeline.