https://hevy.chrisdoc.dev/mcp is the fastest way to start — nothing to install, nothing to keep running, just point your client at the URL and send your API key as a bearer token. The local stdio path runs the server as a child process on your own machine via npx, giving you full local control and working with clients that cannot attach a fixed authorization header to remote requests.
The local
npx path requires Node.js 20 or newer. Run node --version to confirm before proceeding. The hosted endpoint has no local dependency requirements.1
Get your Hevy API key
Open the Hevy app, navigate to Settings → Account, and generate an API key. Copy it and store it somewhere secure — treat it like a password.API access requires an active Hevy PRO subscription. If you do not see the API key option in settings, confirm that your PRO subscription is active.You will supply this key to your MCP client in the next step. Never put it in source control, a URL, a log file, or a CLI argument.
2
Choose your connection method
- Hosted Endpoint
- Local (npx)
The hosted Cloudflare Worker is live at The bearer value is your Hevy API key, not an OAuth token. The Worker validates it with Hevy on each request, does not store it, and forwards it upstream only as Hevy’s required
https://hevy.chrisdoc.dev/mcp and uses Streamable HTTP. Your client must be able to attach a fixed Authorization header to every MCP request.Add the following to your MCP client’s server configuration, replacing your-hevy-api-key with the key you copied in step 1:api-key header.After saving your client configuration, restart or reconnect the client to pick up the new server.3
Ask your first question
Once your client has reconnected and the
hevy server appears in its MCP server list, try one of these prompts:- “Give me a training summary for the last four weeks.”
- “What routines do I have saved on Hevy?”
- “Show my three most recent workouts.”
- “Find exercise templates containing squat.”
- “Which Hevy account is connected?”