Anthropic AI: Claude Fable 5 & Mythos 5 Launched

Anthropic has launched new AI models, Claude Fable 5 and Claude Mythos 5. These models can operate autonomously for longer periods than previous versions. Claude Fable 5 is Anthropic's most powerful model for general use. It excels in programming, data work, document analysis, and scientific research. However, it will not fulfill requests in certain cybersecurity, biology, or chemistry areas. If a threat is detected, requests route to a less powerful model. Claude Mythos 5 shares the same architecture but has fewer cybersecurity restrictions. It will initially be available through Project Glasswing, a partnership with the U.S. government and cyber defense organizations. During testing, Fable 5 migrated 50 million lines of Ruby code in one day, a task that would have taken an engineering team over two months. Mythos 5 was also used in drug discovery research. The cost for using these models is $10 per million input tokens and $50 per million output tokens. For businesses and developers, these new AI tools could significantly impact efficiency and problem-solving.