Anthropic releases Claude Opus 5 with stronger benchmark results
Claude Opus 5 delivers near-flagship AI for coding, search, and knowledge work Get stronger performance at a lower price with Anthropic's most aligned model
Anthropic has released Claude Opus 5 across its apps, Claude Code, and API. The company says the model is designed to be more thoughtful and proactive, while aiming to deliver nearflagship performance at a lower price than its top tier.
According to Anthropic, Opus 5 performs strongly on agentic coding, knowledge work, search, and computer use tasks. The company also says it reached 30.2% on ARCAGI3, ranked first on Artificial Analysis’ Intelligence Index, and scored 42 out of 42 on International Math Olympiad 2026 problems in its internal testing.
Anthropic also described Opus 5 as its most aligned model so far, saying it matches its stronger model on bug finding but remains less capable at writing exploits. The release is positioned as a way to bring higherend capability to users at prices Anthropic says are lower than its premium offering.