Anthropic Launches Claude Fable 5.2 and Mythos 5.2 Models

Focus on hard engineering and science

Anthropic describes Claude Fable 5.2 as particularly good at difficult software engineering problems. It can orchestrate teams of agents and handle long-running projects. The company also says the model has solved some complex scientific problems, including two open problems in theoretical physics.

During prerelease testing, the two models reportedly helped Anthropic's engineers complete "otherwise-infeasible software migrations", design three working electronic devices from scratch, and solve two open problems in the physics of black holes. Anthropic adds that they made substantial progress on robotics evaluations and formalized the proof of Fermat's Last Theorem in 30 hours.

What it means

The claims point to a push toward models that can work autonomously on extended, multi-step tasks rather than answer single prompts. For developers and research teams, agent orchestration and long-running project support are the features most likely to matter in practice. These results come from Anthropic's own testing, and independent verification has not been described.

Access is the other notable detail. Fable 5.2 is open to users now, but Mythos 5.2 will initially go only to vetted organizations, suggesting a more controlled rollout for that model. Anthropic has not said in the material provided when broader access to Mythos 5.2 might follow.