gymcoach
Self-hosted workout tracker with an AI coach that knows your training. Bring your own LLM (Claude or any OpenRouter model). Strength + cardio, offline PWA, one-
What is gymcoach?
Self-hosted workout tracker with an AI coach that knows your training. Bring your own LLM (Claude or any OpenRouter model). Strength + cardio, offline PWA, one-command Docker install. A free, open-source alternative to Strong, Hevy and Fitbod.
Overview of gymcoach
What it is
GymCoach is a self-hosted workout tracker with an AI coach that knows your training. It uses your own LLM key (Claude or any OpenRouter model) for coaching. It is free, open source (MIT license), and a no-subscription alternative to Strong, Hevy, and Fitbod. It supports strength and cardio tracking, works as an offline-first progressive web app (PWA), and can be installed with a single Docker command.
Key features
- AI coach that reads your sessions and provides coaching, using your own LLM key (Claude or any OpenRouter model).
- Fast set logging with shorthand entry (e.g.,
100x8@9), rest timer, plate-loading calculator, and warm-up ramp calculator. - Offline-first PWA for logging without internet connectivity.
- Cardio import from watch files (FIT, TCX, GPX) without requiring a cloud account.
- Established training programs as written (5/3/1, GZCLP, nSuns, PPL).
- Full export of data; nothing is paywalled.
- ChatGPT / MCP connector for external AI integration.
Use cases
- Self-hosted alternative to cloud workout trackers for users who want data privacy.
- AI coaching without a subscription fee.
- Offline logging in environments without internet.
- Importing cardio data from watch files.
- Following established training programs.
Getting started
All you need is Docker on an x86_64 host. Run the following commands to set up:
mkdir gymcoach && cd gymcoach
curl -fsSLo docker-compose.yml \
https://raw.githubusercontent.com/gymcoach-app/gymcoach/main/docker-compose.selfhost.yml
printf 'JWT_SECRET=%s\nPOSTGRES_PASSWORD=%s\n' \
"$(openssl rand -base64 48)" "$(openssl rand -hex 16)" > .env
docker compose up -d
Open http://localhost:3000 and create your account. To enable the AI coach, add ANTHROPIC_API_KEY=... (or LLM_PROVIDER=openrouter with OPENROUTER_API_KEY and OPENROUTER_MODEL) to .env and run docker compose up -d again.
License and pricing
MIT license. Open source. Free. No subscription.
Based on the project's public README and repository metadata.