Anthropic Cuts Agent Costs 75% With New Claude Models

Anthropic has launched Claude Fable 5.1 and Mythos 5.1, its latest and most powerful large language models. The two names reference the same underlying model, with Fable 5.1 as the standard release and Mythos 5.1 available to vetted cybersecurity and life-sciences organizations requiring fewer safeguard restrictions. Beyond benchmark improvements, the release targets enterprise economics by cutting cached context costs by 75%, making persistent AI agents significantly cheaper to run. Anthropic also introduced Enterprise Frontier Safeguards, a new security architecture allowing organizations to retain monitoring data within their own infrastructure.