Anthropic Launches Claude Fable 5.1 and Mythos 5.1 to Advance AI Coding and Scientific Research

Artificial intelligence research firm Anthropic has officially launched its newest frontier AI models, Claude Fable 5.1 and Claude Mythos 5.1. Designed primarily to handle complex computer programming, technical workflows, and advanced scientific reasoning, the twin models represent a significant step forward in autonomous agentic capabilities. While Fable 5.1 is generally available to enterprise clients and developers across major cloud platforms including AWS, Google Cloud, and Microsoft Azure, Mythos 5.1 is restricted to vetted partner organizations working in cybersecurity and life sciences due to its specialized safeguard configurations.

A major highlight of the release is a substantial adjustment to operating costs for enterprise deployments. Anthropic has cut pricing for cached context reads by 75 percent, which reduces overall expenses by up to 45 percent for long-running AI agent tasks. In benchmark evaluations, Claude Fable 5.1 demonstrated impressive performance gains, scoring 52.6 percent on end-to-end scientific research simulations—more than doubling the performance of its predecessor, Fable 5.

Furthermore, Anthropic introduced a refined security framework known as Enterprise Frontier Safeguards (EFS). This framework allows enterprises to store and inspect operational log data entirely within their own cloud infrastructure rather than relying on vendor servers, resolving critical data privacy concerns for sensitive industries. Additionally, the update addresses developer feedback regarding overly cautious safety guardrails, significantly reducing false-positive refusals during cybersecurity analysis and code inspection. As major tech companies compete fiercely in agentic AI capabilities, Anthropic’s latest updates signal a major shift toward making autonomous systems cheaper, more reliable, and easier to govern at scale.