Anthropic has introduced two new artificial intelligence models, Claude Fable 5.2 and Claude Mythos 5.2, which the company says are built to handle especially difficult technical work. The announcement positions Claude Fable 5.2 as the more broadly available option, while Claude Mythos 5.2 will initially be limited to vetted organizations.
According to Anthropic, both models deliver gains in coding, reasoning, and vision compared with their 5.1 predecessors, and they can produce outputs more quickly. The company describes the new versions as suited to what it calls daunting technical challenges.
What Anthropic Says the Models Can Do

Anthropic highlights Claude Fable 5.2’s strength in demanding software engineering tasks. The model can coordinate teams of agents and manage projects that run over extended periods, according to the company. It has also been credited with solving complex scientific problems, including two open theoretical physics questions.
During prerelease testing, Anthropic says the models helped its engineers carry out software migrations that would otherwise have been infeasible. The company also claims the models designed three working electronic devices from scratch, addressed two open problems in black hole physics, made substantial progress on robotics evaluations, and formalized a proof of Fermat’s Last Theorem in 30 hours.
Those claims are notable for their scope, though Anthropic has not provided independent verification in the announcement. The company presents them as examples of the models’ capacity for long-running, technically demanding work.
Availability and Prior Releases

Claude Fable 5.2 is available starting today. Claude Mythos 5.2, by contrast, is being released only to vetted organizations at the outset. The two models follow the September release of Claude Fable 5.1 and Claude Mythos 5.1.
The staggered rollout suggests Anthropic is taking a more cautious approach with Mythos 5.2, though the company has not detailed the criteria for vetting or the reasons for the restricted access. Fable 5.2’s immediate availability gives a wider set of users access to the company’s latest coding and reasoning capabilities.
Positioning in a Competitive Field

Anthropic’s announcement emphasizes practical engineering and scientific applications rather than general-purpose chat. By pointing to software migrations, device design, and physics proofs, the company is framing the 5.2 models as tools for specialized technical work. The mention of agent orchestration and long-running projects also signals a focus on autonomous, multi-step tasks.
Whether those claims hold up under broader use will become clearer as users put Claude Fable 5.2 through its paces. For now, Anthropic is presenting the new models as a step forward in both capability and speed, with Mythos 5.2 reserved for organizations it deems appropriate for early access.
