For years, the only way to analyze your Strava history with an AI assistant was clumsy: export your data, paste it into a chatbot by hand, and repeat the whole process every time you wanted a fresh answer — or trust an unofficial third-party tool with your account. Strava's new MCP Connector removes that friction with a secure, official connection between your Strava data and Claude.
With it, Strava becomes one of the first major connected-fitness platforms to ship a native AI assistant integration through the Model Context Protocol.
What is the Strava MCP Connector?
The MCP Connector is a remote Model Context Protocol server operated by Strava. MCP is an open standard for connecting AI assistants to external data sources in a structured, permission-scoped way. The connector gives you conversational access to your own activity data: you authorize it once via OAuth, then ask questions in plain language and Claude pulls the relevant data to answer.
As Ryan Dixon, Strava's VP of Partnerships & Developer Relations, put it in the announcement: “Athletes have been telling us, in increasingly creative ways, that they want more ways to analyze their own training data.”
What data can it access?
The connector is read-only and scoped to your individual account. Once connected, Claude can reach:
- Full stream data — per-second metrics such as heart rate and pace
- GPS data — for geographic and route-based analysis
- Power data — for cycling activities
- Club and event data — context beyond individual activities
Read-only by design
The MCP Connector cannot upload, edit, or delete activities. It only reads the data in your account so an AI assistant can help you understand it. Access is scoped to your account and can be revoked at any time from your Strava settings.
Who can use it?
The MCP Connector is available to all Strava subscribers, and Strava began rolling it out globally the week of the June 1, 2026 announcement. At launch it works with Claude, Anthropic's AI assistant.
What can you ask it?
Because Claude can read your streams, GPS, and power data directly, you can ask the kinds of questions that used to require a spreadsheet or a dedicated analytics app:
- “How has my average heart rate at threshold pace changed over the last six months?”
- “Compare my climbing power on my last three rides up the same segment.”
- “Which weeks this year did I increase training load too quickly?”
- “Summarize my longest runs and how my pace held up in the final third.”
How to connect it
Setup runs through an OAuth authorization flow — the same consent screen you see when connecting any app to Strava — so your credentials are never shared with the AI assistant. Strava publishes the current step-by-step instructions in its Help Center:
When you're done, you can revoke the connection at any time from your Strava connected-apps settings, just like any other authorized application.
How it fits Strava's broader AI strategy
The MCP Connector is also a statement about how Strava wants AI to access its data. Strava's API agreement already prohibits third parties from using data obtained through the API to train AI or machine-learning models. By offering an official, permission-scoped MCP server, Strava gives athletes a sanctioned path to AI analysis while keeping control of exactly what gets shared and how. If you're a developer, we cover what this means for building on the platform in our guide to Strava's 2026 developer program changes.
The bottom line
If you're a Strava subscriber, the MCP Connector is the easiest and safest way yet to turn years of activity history into answers. It's read-only, account-scoped, revocable, and official — no manual exports, no copy-paste, and no handing your login to an unofficial tool. Prefer a purpose-built dashboard instead? Browse our directory of analytics and training apps for Strava.