Claude Fable 5.2 and Mythos 5.2: Anthropic's new models explained

Anthropic released Claude Fable 5.2 and Mythos 5.2, with the video covering benchmark performance, real-world capabilities including building an operating system, and early tester reactions. The models reportedly solved open physics problems, completed complex calculations in 30 hours, and demonstrated hardware design abilities.

The video compares performance to other leading models, discusses pricing implications and competitive dynamics in the market, and explores whether these models represent progress toward artificial general intelligence. It also covers Anthropic's decision to hold back certain capabilities.