Anthropic's Opus 5 nearly matches its flagship Fable model at half the price

[Darren51]

Anthropic released Claude Opus 5, describing it as a thoughtful and proactive everyday model that comes close to the frontier intelligence of Fable 5 at half the price. On coding and knowledge work evaluations like Frontier-Bench and GDPval-AA, Opus 5 is the new state of the art among Anthropic's models, though it remains behind Mythos 5 specifically on cybersecurity tasks. It is now the default model on Claude Max and the strongest model available on Claude Pro

The release is Anthropic's fourth Claude 5 series model in under two months, following Opus 4.8, Sonnet 5, Mythos 5 and Fable 5, a pace Axios described as evidence that AI deployment has shifted from blockbuster launches to rapid, continuous improvement on capability, cost and speed. On ARC-AGI 3, a benchmark testing novel problem solving, Opus 5 scored three times as high as the next best model, and on the OSWorld 2.0 computer use benchmark it surpassed Fable 5's best result at just over a third of the cost

Anthropic says its pre deployment automated behavioral audit found Opus 5 to be its most aligned model to date, with the lowest rates of deceptive behavior and the least susceptibility to being tricked into misuse. Pricing stays the same as its predecessor Opus 4.8 at 5 dollars per million input tokens and 25 dollars per million output tokens, and the launch includes a new opt in Automatic Fallbacks feature that reroutes any request blocked by a safety classifier to another model instead of simply returning a refusal

Introducing Claude Opus 5www.anthropic.com

[Ellie85]

Three times the next best model on ARC-AGI 3 is a genuinely huge gap for a benchmark specifically designed to resist memorization

[Wrench]

Four Claude 5 series releases in under two months is a pace that would have been unthinkable for a single lab just a couple years ago

[NatureBoyOwen16]

The wind tunnel and cell artifact demos are a nice touch, visual output quality rarely gets its own section in a model announcement like this