Introducing Claude Fable 5.2 and Claude Mythos 5.2

Anthropic released Claude Fable 5.2 and Claude Mythos 5.2, an upgraded pair built on the same underlying model, with Fable carrying cybersecurity and biology safeguards for general availability while Mythos is reserved for vetted life-sciences and cyber researchers. The models generate outputs up to 30% faster than versions 5.1 and are priced lower at $8 per million input tokens and $40 per million output tokens, with an optional fast mode at $16 and $80 that speeds token generation by up to 2.5 times. Anthropic said prerelease testers used the models to complete otherwise-infeasible software migrations, design three working electronic devices from scratch, and formalize a proof of Fermat's Last Theorem in 30 hours. The company reported that misalignment risks remain low, that Fable 5.2's safeguards produce fewer unnecessary blocks than 5.1's, and that it has added its strongest protections yet against distillation attacks. Anthropic positions Fable and Mythos for ambitious open-ended technical work, steering most coding and knowledge tasks to Claude Opus 5.5, with Fable 5.2 also usable as an advisor or orchestrator for smaller models. Next, watch enterprise adoption of the premium tiers and whether the claimed efficiency gains hold under independent benchmarking.