At first, ordinary runners learned to run by reading books. Later, as the running boom arrived, joining running clubs to learn and hiring coaches gradually became common.

In these two years of the AI era, there seems to be a new way of doing things. However, whether running apps or running social platforms, their original infrastructure designs had little to do with AI.

For example, if a runner wants toAI to analyzethe activities on Stravadata, the only way is to exportCSV the file, then copy and paste it into an AI chat box, and when there's a new question next time you have to repeat the whole process — very cumbersome.

But from6 month1 day, Strava officially connected Claude large model, and the previous compromise solutions will become history.

Strava officially announcedthat it launchedMCP a connector that can connect AI large modelsClaude seamlessly into user accounts.

After runners authorize, they can use simple, straightforward phrasing toask Claude various questions about their own training, and the AI will draw the most accurate analysis from the runner's dynamic real-time data.

Whether it's heart rate, pace,GPS tracks, or brandAlong with race history, the AI will provide a complete "panorama" rather than an outdated, static snapshot.