← Back to roadmap

Space Statistics

Every space now has a Stats tab that aggregates invocation data across its cases: turn count, input and output tokens, average tokens per turn, average duration, and breakdowns by model and agent. The same breakdown is available at the case level, so you can drill from a whole space down to a single case.

Stats reflect invocations from the feature’s release onward; historical activity is not backfilled. Operational LLM calls like title and summary generation aren’t counted yet.