Strava has announced a new AI-powered feature that allows subscribers to connect their accounts directly to Claude, the AI assistant developed by Anthropic.
The new Model Context Protocol (MCP) connector is rolling out globally this week and enables you to ask questions about your training history in natural language.
According to Strava, the integration can access activity data, including heart rate, pace, GPS data, cycling power metrics, clubs, and events.
The launch makes Strava one of the first major fitness platforms to offer a native integration with an external AI assistant, giving subscribers a new way to explore the data they've accumulated over months or even years of training.
A different role to Athlete Intelligence
The announcement arrives less than two years after Strava introduced Athlete Intelligence, its AI-powered feature that generates summaries and insights for individual activities.
While both features use AI, they are designed for different purposes.
Athlete Intelligence focuses on explaining what happened during a specific workout, highlighting notable efforts, pacing trends and other key takeaways after an activity has been completed.
The new Claude integration is more conversational, allowing you to ask your own questions about your training history and receive responses based on data stored in your Strava account.
Strava suggests subscribers could ask questions such as "Are my easy days easy enough?" or "How is my cross-training affecting my running?"
Making training data easier to explore
According to Strava, athletes have already been using exports, spreadsheets and third-party tools to analyse their training history with AI assistants.
The MCP connector is intended to simplify that process by creating a direct connection between Strava and Claude.
The company says the integration provides read-only access to user data and can be revoked at any time through Strava settings.
Ryan Dixon, Vice President of Partnerships & Developer Relations at Strava, said athletes have been looking for new ways to analyse their training data for years.
"The MCP connector gives them a far more efficient, safer tool while keeping the athlete in control," Dixon said.
As with any AI integration that accesses personal data, you will likely want to review the permissions involved before connecting your account, particularly given the amount of location, performance and health-related information stored within Strava histories.
The Strava MCP connector is available to subscribers and is now rolling out globally. Setup instructions are available through Strava's Help Centre.
