Anthropic has released Claude Opus 5.5, the latest addition to its AI model portfolio, which it says matches the performance of the company’s Claude Fable 5.1 model.
The company also claims a 40% reduction in operating costs compared to previous versions.
Go deeper with GlobalData
Access deeper industry intelligence
Experience unmatched clarity with a single platform that combines unique data, AI, and human expertise.
Claude Opus 5.5 is now available on platforms such as Amazon Web Services (AWS), Microsoft Azure, and Google Cloud.
According to Anthropic, Claude Opus 5.5 is the inaugural model in the Claude 5.5 family and is the company’s first launch following its recent call for caution in advancing high-powered AI technologies.
The model is said to have undergone external testing by evaluators, including Frontier Design and METR, prior to release.
Anthropic states that Opus 5.5 achieved the best results to date on its automated behavioural alignment audits, which simulate thousands of scenarios to measure safety and adherence to assigned boundaries.
The company reports that Opus 5.5 brings significant advances over the earlier Opus 5 model, particularly in handling complex tasks.
Early testers completed tasks such as a 680,000-line code migration within a day, an activity the company says would have typically required weeks for a full engineering team.
The model is also described as adept at software efficiency evaluations and demonstrated improved output in creative tasks, according to independent tests cited by Anthropic.
In terms of cybersecurity and biology, Opus 5.5 reportedly performs on par with Claude Mythos 5.1, prompting the company to apply similar safety safeguards found in its highest-performing models.
Operationally, Anthropic reports that the model requires less computing power to serve, providing input at $4 per million tokens and output at $20 per million tokens, which is 20% below Opus 5’s pricing.
Cache reads are listed at $0.20 per million tokens, equivalent to a 60% reduction in cost, while generation speed is said to be over 30% faster than Opus 5.
The company has also increased usage limits for several subscription tiers and introduced a new rate limit reset feature.
Anthropic emphasised improved communication capabilities in Claude Opus 5.5, with early testers finding its outputs clearer and more structured than those of previous models.
Additional models in the Claude 5.5 family, including Claude Sonnet 5.5 and Claude Haiku 5.5, are expected to be released in the coming weeks.
Recently, Anthropic and Accenture announced a partnership to embed teams within Anthropic for the evaluation and assessment of advanced AI model safety. Both companies plan to invest at least $1bn each over the next five years to support capacity building in AI safety.
