Anthropic has launched the newest model of its largest Claude mannequin, Opus. Claude Opus 4.5 is best at dealing with complicated reasoning than earlier Claude fashions and makes enhancements throughout agentic device use, pc use, novel downside fixing, and extra.
The corporate says early testers of the brand new mannequin declare that it handles ambiguity higher and causes over tradeoffs with no need human intervention. “They informed us that, when pointed at a posh, multi-system bug, Opus 4.5 figures out the repair. They stated that duties that had been near-impossible for Sonnet 4.5 only a few weeks in the past are actually inside attain. General, our testers informed us that Opus 4.5 simply ‘will get it,’” Anthropic wrote in a put up.
This launch additionally coincides with a brand new effort parameter being launched within the Claude API, permitting builders to resolve how a lot effort Claude ought to spend on an issue. In response to Anthropic, Opus 4.5 makes use of considerably fewer tokens than its predecessors to unravel issues, even at its highest effort degree. For instance, at a medium effort degree, Opus 4.5 matches Sonnet 4.5’s rating on SWE-bench Verified whereas utilizing 76% fewer output tokens, whereas on the highest effort degree it makes use of 48% fewer tokens whereas exceeding Sonnet 4.5’s efficiency by 4.3%.
When it comes to security, Anthropic claims that Opus 4.5 Pondering is much less inclined to immediate injection assaults than fashions like Sonnet 4.5 Pondering, GPT-5.1 Pondering, or Gemini 3 Professional Pondering.
Anthropic additionally introduced updates to Claude Code that coincide with the brand new mannequin’s launch. First, Plan Mode can create extra exact plans and execute them extra totally, with Claude asking clarifying questions upfront and constructing them into the plan earlier than executing it. Second, Claude Code is now accessible in Anthropic’s desktop app, which permits a number of native and distant classes to be run side-by-side.
Moreover, the Claude app will now routinely summarize earlier elements of an extended dialog, and the Claude for Excel beta is increasing to Max, Group, and Enterprise customers.
Opus 4.5 is obtainable now throughout all of Anthropic’s apps and API and prices $5 per million enter tokens and $25 per million output tokens.
“Opus 4.5 is a step ahead in what AI methods can do, and a preview of bigger adjustments to how work will get achieved,” the corporate wrote.
