Anthropic says during prerelease testing Mythos 5.2 and Fable 5.2 formalized the proof of Fermat's Last Theorem and solved two open black hole physics problems (Jay Peters/The Verge)

Anthropic has released Mythos 5.2 and Fable 5.2, which the company describes as its latest best-in-class models. According to The Verge, during prerelease testing, these models formalized the proof of Fermat’s Last Theorem and solved two open problems in black hole physics. TechCrunch reports that the new models show improvements in reasoning, vision, and coding, while communicating more clearly with users.

The update includes performance and pricing changes for the Fable model. TechCrunch notes that Fable 5.2 will return responses 30% faster across effort levels, with faster inference serving available for double the price. 9to5Mac reports that the models generate outputs up to 30% faster than the 5.1 versions. Anthropic is also lowering the cost of Fable to $8 per million input tokens and $40 per million output tokens, making it cheaper than GPT-6 Astra, according to TechCrunch.

Fable and Mythos are designed for ambitious, open-ended technical work, while Anthropic recommends Claude Opus 5.5 for most daily coding and knowledge tasks, according to 9to5Mac. Additionally, TechCrunch reports that Fable 5.2 is compatible with zero data retention policies via the Enterprise Frontier Safeguards system.

Brief written by urgent.news from Techmeme, TechCrunch, 9to5Mac — 3 reports on this story. Machine-written — may contain errors; check the original before relying on it.