Anthropic ships Claude Fable 5.2 and Mythos 5.2 days before its $1T-plus IPO
What's the deal? Anthropic released Claude Fable 5.2 and Claude Mythos 5.2 on Monday, upgraded models built for demanding reasoning and long-running agentic work. Fable 5.2 is publicly available, while Mythos 5.2 stays limited to organisations vetted through its Life Sciences and Cyber Verification programs.
The specs: The models generate outputs up to 30% faster than the 5.1 generation and cost about 20% less, at $8 per million input tokens and $40 per million output tokens. An optional fast mode speeds up token generation by up to 2.5 times for $16 and $80 per million tokens.
The claims: Anthropic says that during prerelease testing, the models solved two open problems in the physics of black holes, designed three working electronic devices, and formalised the proof of Fermat's Last Theorem in 30 hours. The Fermat result was a computer-checked formalisation using the Lean tool by a team of AI agents, not a new proof of the theorem, and the physics results remain the company's own claims pending outside validation.
Why now? The launch is the first frontier release since chief executive officer Dario Amodei called in September for a slowdown in the pace of AI development. Anthropic's head of product management for research, Dianne Penn, told CNBCDealroom has a profile for this one. Try Dealroom → the company had called for "defining the pace" of AI development rather than a full pause.
The IPO: The release lands as Anthropic prepares for an IPO that CNBC reports could come as early as next week, with the company seeking a valuation between $1 trillion and $1.5 trillion. In an amended prospectus, Anthropic said revenue for the nine months ended September 30 was $33.24 billion, roughly 1,200% more than a year earlier.
Market positioning: Anthropic says Fable and Mythos are best for the most ambitious open-ended technical work in engineering, mathematics, and robotics; it recommends Claude Opus 5.5 for most daily coding and knowledge tasks. The company says over 90% of code merged into its core codebase is now written by Claude.
The signal: The release doubles as a high-profile demonstration of capability just ahead of the roadshow, in a market where Google and OpenAI have also recently shipped new frontier models. By pairing lower operating costs with bold scientific claims, Anthropic is making its case for a trillion-dollar valuation, even as its headline breakthroughs await independent verification.
Read more: anthropic.com, The Verge
Chart: Ethan MollickDealroom has a profile for this one. Try Dealroom →. Claude Fable 5.2 versus existing trendlines on the Remote Labor Index, Epoch Capabilities Index and RoboDojo. Data as of 10 November 2026.
