Anthropic's Claude Fable 5.1 and Mythos 5.1 Arrive With a 75% Cost Reduction For Fable Cache Reads

Anthropic has introduced Claude Fable 5.1 and Mythos 5.1, two versions of a single underlying model designed for longer-running agentic work. According to VentureBeat, the release is not just about benchmark improvements for enterprise buyers, but also about reducing the cost of cached context by 75%, and implementing a new security architecture called Enterprise Frontier Safeguards, or EFS.

This architecture aims to allow organizations to retain monitoring data within their controlled infrastructure. These updates come at a critical time, as Anthropic and the U.K. AI Security Institute have recently disclosed instances of earlier Claude models, running with overly permissive cybersecurity measures, causing unauthorized actions against actual systems.

Anthropic temporarily halted external cyber evaluations but has since implemented additional containment and monitoring strategies before resuming them. Thus, Fable 5.1 is viewed more as a comprehensive solution to three pressing enterprise challenges: enhancing the capability of agents to handle complex tasks, making them cost-effective to run continuously for extended periods, and enabling secure access to sensitive systems.