Anthropic Launches Claude Opus 5.5, Its Most Powerful Model Yet, At 40% Lower Cost

Anthropic has unveiled Claude Opus 5.5, its most powerful AI model yet, boasting a 40% reduction in cost compared to its predecessor, Opus 5. The new 5.5 model is positioned as the company's top-tier AI system, demonstrating performance comparable to Claude Fable 5.1, currently Anthropic's leading model, while being significantly more affordable.

Despite sitting one tier below Fable 5.1, Opus 5.5 outperforms it on benchmarks for agentic coding, knowledge work, computer usage, visual chart recognition, and multidisciplinary reasoning.

Anthropic highlights improvements in both performance and safety with Opus 5.5, noting faster output generation—more than 30% quicker than the previous model—and more natural communication, with key information appearing earlier in responses for easier scanning. Pricing adjustments include a 20% reduction in per-token costs, bringing typical workload expenses down by around 40%.

Additionally, cache-read pricing has dropped by 60% to $0.20 per million tokens, with an option for a faster processing mode at a premium rate of $8 per million input tokens and $40 per million output tokens.

Safety has been enhanced with Opus 5.5, mirroring the protective measures of Claude Mythos 5.1, Anthropic's most secure system. This includes restricting the model's capabilities for cybersecurity exploitation or biological weapon development, though organizations requiring these features for legitimate research can apply through Anthropic's Life Sciences Verification Program.

A new safeguard, "preserved thinking," prevents bad actors from extracting the model's reasoning through fake accounts. Opus 5.5 underwent testing by external evaluators such as Frontier Design and METR before its release, scoring highest among Anthropic's models on the company's internal behavioral audit.

Available immediately across all Anthropic platforms, including Claude Platform, Amazon Web Services, Google Cloud, and Microsoft Azure, Opus 5.5 comes with extended five-hour usage limits for subscribers on Pro, Max, Team, and Enterprise plans. The company has also reset rate limits for usage, effective until October 22. Upcoming models, Claude Sonnet 5.5 and Claude Haiku 5.5, are expected to follow with further performance and efficiency enhancements.