Anthropic has released Claude Fable 5.2 and Claude Mythos 5.2, delivering significant performance and cost improvements. Both models generate responses up to 30% faster than their 5.1 predecessors, while pricing has dropped to $8 per million input tokens and $40 per million output tokens.
On the FrontierCode 2.0 programming benchmark, Fable 5.2 achieved 49.1% on the Xhigh effort level—the highest score among all evaluated systems. By comparison, GPT-6 Astra, operating at High reasoning level, scored 42.8%. Fable 5.2 is now available across all paid Claude plans, though Mythos access remains restricted.