Anthropic has launched two new versions of Claude—Fable 5.1 and Mythos 5.1—which use the same underlying model but differ in their safety guardrails. The distinction allows the company to offer tailored versions for different use cases and user groups.
The new models deliver substantial cost savings. Standard token-based tasks see approximately 25% price reductions, while agentic-heavy workloads can achieve savings of up to 45%. For cybersecurity applications, false positive rates have dropped by 60%, making these versions particularly valuable for security teams managing high-volume threat detection.
Fable 5.1 is available to the general public immediately, while Mythos 5.1 remains restricted to selected groups. These releases signal Anthropic's focus on making capable AI more accessible and cost-effective without compromising performance or safety.