Anthropic clarifies Claude Code effort display bug: no performance impact

trq212 · x · 2026-08-23

An Anthropic employee clarified that the 'High' reasoning effort value showing as 10/100 in Claude Code is due to a test configuration mapping, not a model downgrade. Official evaluations confirm actual output quality is unaffected. Users experiencing regressions are encouraged to report issues for credits.

Original post →

More from Companies & People

Companies & People channel →