Anthropic released an updated model Claude Opus 5.5

Company Anthropic introduced the updated language model Claude Opus 5.5. The release is aimed at increasing the efficiency of agent-based computations and optimizing costs while maintaining high performance in complex tasks.

The new version demonstrates a result of 252 points in performance tests, completed in 1 hour. Developers claim that the cost of using the model has decreased by 40% compared to the previous iteration of Opus 5. At the same time, the new model is comparable in its functional capabilities to the Claude Fable 5.1 model.

Key improvements affected three areas: agent programming, interaction with computer interfaces (computer use), and work with knowledge bases. Engineers implemented updated security protocols and algorithms that provide more natural communication with the user.

The technical characteristics of the model allow it to be used in scenarios that require high accuracy when executing multi-step instructions. The improved architecture minimizes errors when processing long contexts, which is critical for automating business processes.

The update continues the company's strategy to lower the entry threshold for corporate clients. Detailed discussions of technical aspects and initial community feedback are available on the Hacker News platform.