Opus 5 targets stronger everyday professional work

Anthropic introduced Claude Opus 5, positioning it near Fable 5 on many coding and knowledge-work evaluations at a lower price. The announcement describes different reasoning-effort settings and makes Opus 5 the default on Max and the strongest available option on Pro at launch.

Context

Reported performance is most useful when read alongside the effort setting and cost of the task. The model’s practical value depends on the work users actually repeat. A stable comparison should include errors and review burden, rather than assuming the strongest benchmark configuration is the best everyday default.