Plans and routines, tracked set by set.
Build weekly plans from routines, add rest days, and track planned sessions. Log reps, load, RIR, supersets, and rest periods set by set.
Plans, routines, workout history Learn more
Feature Highlights
Built for serious tracking without the noise.
Train Libre keeps the daily surface calm and the underlying data rich, so you can log precisely, review honestly, and understand progress over time.
Nutrition, hydration, and supplements together.
Track meals, calories, macros, fluids, caffeine, creatine, and custom supplements in one local diary.
Food, water, macros, dosesBody data with context.
Log bodyweight and measurements, then read them beside nutrition trends and training consistency.
Measurements, trends, goalsRecovery, sleep, steps, and pulse insights.
Use sleep, steps, heart-rate context, and recovery views where your device data is available.
Learn more Training guidance, not diagnosisImport, back up, and share your data.
Create local backups, import and export your data, and use one-way health export without making the cloud your source of truth.
Local-first controlAI meal recognition that stays reviewable.
Capture meals from photos or text, then review portions, food matches, warnings, and locally computed nutrition before saving.
Learn more Photo, text, grams, review
AI, nutrition & training models
Useful estimates, with clear limits.
AI capture and training or nutrition estimates solve different problems. Each has its own inputs, limits, and review flow.
From plate to log, with a checkpoint.
Describe or photograph a meal. Your selected AI provider proposes foods and portions; Train Libre matches them to catalog foods so you can review and edit the result before saving.
Example nutrition totals
800 kcal
P 40g
C 116g
F 20g
Photo and text capture
Suggested portions
Catalog-based nutrition
Review before saving
A trend-informed estimate. Targets stay yours.
The nutrition overview keeps daily intake in view. A separate weekly review shows goal progress and lets you decide whether to apply a suggested target change.
How the models work
Algorithmic transparency
These summaries describe what each feature uses and returns. The linked articles document implementation details and limits.
AI meal capture
AI proposes foods and portions; the app matches them to catalog foods for your review.
- Nutrition totals use matched food data.
- Review and edit suggestions before saving.
Maintenance estimate
A recursive estimate combines logged intake with a smoothed weight trend and a profile-based starting prior.
- Uses a 14-day input window.
- Sparse data can prevent an observation update.
Weekly nutrition goal review
A review compares logged progress with the active weight goal and may propose next steps.
- Data and compatibility checks gate adjustments.
- Applying a recommendation requires user action.
Macronutrient targets
A deterministic allocator divides a calorie target into protein, fat, and carbohydrate targets.
- Uses body weight and the selected goal.
- Fixed app rules, not an individualized prescription.
Workout progression
When enabled, the workout can suggest a starting target from the latest comparable session.
- Off by default; suggestions can be edited.
- Only eligible load-and-repetition sets receive targets.
Estimated one-rep max
A logged load and repetition count produce a formula-based strength estimate.
- Uses the Brzycki equation within its valid repetition range.
- An estimate, not a measured maximum or safety assessment.
Muscle recovery estimate
Eligible completed sets contribute to a time-decaying load estimate for mapped muscles.
- Uses set details and exercise-to-muscle mappings.
- A training-load heuristic, not a measure of whole-body recovery.
Sleep Health Score (SHS v3)
A 0–100 heuristic combines available sleep duration, continuity, stages, timing, and regularity.
- Missing inputs affect score completeness.
- Not a validated clinical score or diagnosis.
Privacy & Ownership
Your data starts local.
Train Libre is offline-first. Your tracking data is handled locally by default, and there is no mandatory Train Libre cloud account. Sharing, export, health integrations, catalog refreshes, AI meal recognition, and provider calls happen only through features you choose to use.
Local by default
Workouts, meals, measurements, and app state are designed around device-local storage.
Portable when needed
Backups, imports, exports, and share sheets make data movable without turning it into a hosted account.
Clear boundaries
Optional AI and health features are separate choices, with your API key and platform permissions in your control.
Product Showcase
The actual app, framed with care.
Dark surfaces, bold typography, glass controls, and restrained color accents carry the same visual language from logging to AI capture to training review.
Open & Transparent
Public code. Public catalogs. Private logs.
Train Libre is available on GitHub and built around understandable data flows rather than opaque tracking loops.
Developed in public
The project source is available on GitHub, with app behavior, catalog tooling, backup logic, and privacy boundaries visible in the repository.
Open catalog integrations
Food and exercise coverage is grounded in public sources, including Open Food Facts and wger-based catalog data.
Transparent analytics
Progress, recovery, consistency, and nutrition insights are meant to be readable, practical, and grounded in the data you logged.