Anthropic released Claude Opus 5.5 on Sept. 22 at $4 per million input tokens and $20 per million output tokens, down from $5 and $25 for Opus 5. Cache reads cost $0.20 per million tokens, a 60% reduction. The new model requires less computing power to serve and generates output more than 30% faster. Anthropic's tests put performance at the level of Claude Fable 5.1 on most tasks, with typical workloads costing 40% less than Opus 5 at default settings.

Savings vary with usage and reasoning settings. At maximum effort, Opus 5.5 scored 58 on the Artificial Analysis Intelligence Index, the highest result measured, but used about 1.6 times as many output tokens per task as Opus 5. Its cost per task in those tests was $5.98, compared with $5.86 for Opus 5, as increased usage offset the lower prices.

Opus 5.5 is now the default in Claude Code and the Claude app, including Cowork, for Pro, Max and Team plans. Claude Code's five-hour session limits rose 20%, and subscribers received a usage reset they can apply through Oct. 22. The model is also available through AWS, Google Cloud, Microsoft Azure, Cursor and GitHub Copilot. It adds Fable 5.1-class safeguards for cyber, biological and frontier AI development requests, routing flagged requests to another model.