If you’re a subscriber you can now pick Claude Sonnet 5.5 in the Chat as an alternative to Gemini 3.1 Pro.
Why we’re adding it: Anthropic released Sonnet 5.5 yesterday, and we wanted to give you a second strong option for the heavier work: rebuilding a week around a change of plans, comparing recent sessions, reasoning through recovery and race prep.
What our testing showed: on our internal test suite it handled workout creation, skipping and rebuilding, and activity analysis as reliably as Gemini 3.1 Pro, with no failed tool calls in the live run. On the same questions it answered at least as fast as Gemini 3.7 Flash and faster than Pro, and it was quickest on analysis questions that pull several activities together.
How is it different to Gemini 3.1 Pro: From our limited testing, Claude seems to be more literal and more careful. It checks before it would duplicate a workout, it tells you when data is missing rather than estimating, and it asks when a request could mean two things. Gemini Pro tends to be more decisive on it’s own.
Beta: If a reply looks wrong, an approval doesn’t go through, a workout isn’t what you asked for, or anything else goes wrong or feels off please reply here with roughly a description of what happened.
