As we speak we’re releasing Claude Opus 4.1, an improve to Claude Opus 4 on agentic duties, real-world coding, and reasoning. We plan to launch considerably bigger enhancements to our fashions within the coming weeks.
Opus 4.1 is now obtainable to paid Claude customers and in Claude Code. It is also on our API, Amazon Bedrock, and Google Cloud’s Vertex AI. Pricing is similar as Opus 4.
Claude Opus 4.1
Opus 4.1 advances our state-of-the-art coding efficiency to 74.5% on SWE-bench Verified. It additionally improves Claude’s in-depth analysis and information evaluation abilities, particularly round element monitoring and agentic search.
GitHub notes that Claude Opus 4.1 improves throughout most capabilities relative to Opus 4, with significantly notable efficiency features in multi-file code refactoring. Rakuten Group finds that Opus 4.1 excels at pinpointing precise corrections inside giant codebases with out making pointless changes or introducing bugs, with their group preferring this precision for on a regular basis debugging duties. Windsurf reviews Opus 4.1 delivers a one customary deviation enchancment over Opus 4 on their junior developer benchmark, exhibiting roughly the identical efficiency leap because the leap from Sonnet 3.7 to Sonnet 4.
Getting began
We advocate upgrading from Opus 4 to Opus 4.1 for all makes use of. When you’re a developer, merely use claude-opus-4-1-20250805 through the API. You can too discover our system card, mannequin web page, pricing web page, and docs to be taught extra.
As all the time, your suggestions helps us enhance, particularly as we proceed to launch new and extra succesful fashions.
