Gemini can now analyse your Strava runs and find why your pace slipped

What the new Gemini Strava integration does

Strava has started rolling out an integration with Google's Gemini assistant that lets runners ask questions about their own training history in plain language. According to Android Police, the feature began appearing in the United States on 28 October 2026, and Strava has not said when it will reach other countries.

To use it, runners need a Strava subscription, a US account and any Gemini plan. Once connected, the assistant can read activity data such as pace, heart rate, distance and the weather recorded with each run, then answer questions about trends over weeks or months.

The feature was put to the test by Android Police writer Ben Khalesi, who has covered AI and Android since 2023 and trained as an AI engineer. His verdict: it works best as a quick way to search your fitness history, but you still need to check its answers against the actual activities.

Consistency, not excuses, was the real problem

The first test was a simple one: why did familiar neighbourhood runs feel heavy and slow even though heart rate stayed in the usual aerobic zone? Instead of scrolling through three months of activities manually, the writer gave Gemini a date range and asked it to compare pace against effort across a whole training block.

Gemini mapped volume and pace over time, then compared training cycles, focusing on the weeks before the sluggish runs. Its conclusion was blunt. Weekly mileage had fallen from about 15 miles to roughly 8, largely because the number of runs each week had halved. The assistant suggested rebuilding mileage gradually before chasing pace again, since aerobic fitness had dropped along with consistency.

That answer cut through the comfortable excuses the writer had been making to himself about weather and tired legs.

Weather and pacing got partial blame

Not every slow run was down to training habits. When the writer pointed Gemini at the worst run of his month and asked whether heat was the culprit, the assistant used the temperature, humidity, wind and general conditions stored with the activity. That day's humidity was high.

Gemini agreed conditions were tough and estimated a small pace penalty from the heat, but it refused to hand over a free pass. It still named the drop in consistency as the main reason for the slow pace.

A third question looked at pacing within a single bad run. Asked for mile-by-mile splits, Gemini flagged the mile where pace began to drift and set it against an overly fast start, explaining that the first 5K had probably been run too quickly. The writer noted that reviewing split decay this way beats building a spreadsheet.

Cross-sport trends proved the most useful

The most valuable insight came from data outside running. The writer also logs gym sessions, so he asked whether strength training fatigue was behind his heavy legs. Gemini can find trends across all logged sports, and it lined up lifting days against running days.

It found that heavy leg sessions were being scheduled immediately before runs, and that the lack of recovery between them may have contributed to the sluggish feeling. Of everything the assistant surfaced, the writer called this the most useful connection it made.

Who can use it and what to keep in mind

For now, the feature is limited to US-based Strava subscribers who also have a Gemini plan. Strava has not announced a timeline for other regions, so runners elsewhere will need to wait.

There are also practical limits. The assistant needs specific dates and locations to build an accurate history, which means users need some idea of what they are looking for before they ask. Vague questions will not get useful answers.

The writer's advice is measured. He would not rebuild a training plan based on a chatbot summary alone, but he found the feature gave an honest reality check about consistency.