Anthropic has released Claude Fable 5.1 and Claude Mythos 5.1, advancing its frontier coding and knowledge-work models with cost reductions of 25 to 45 percent through aggressive pricing cuts on cached input processing. Fable 5.1 achieves the same or better performance than its June predecessor while delivering improved safety guardrails and reduced false positives for security workflows.
Performance and Cost Structure
Fable 5.1 outperforms Fable 5, Opus 5, and OpenAI’s GPT, 5.6 Sol across multiple coding and reasoning benchmarks. The cost advantage stems from reduced pricing on cache reads, the mechanism that allows models to reuse previously processed context without recomputing it. For organizations running repetitive or multi-turn workflows, this pricing shift fundamentally changes the economics of deploying frontier models.
In testing with investment firm Millennium, Fable 5.1 identified the root cause of a rare system crash that neither Millennium’s own engineers nor competing models had been able to explain. The model avoids shortcuts that produce lower-quality work and instead pursues deeper problem-solving on long-running tasks.
Safety Improvements and Cybersecurity Features
Claude Code users see approximately 60 percent fewer cybersecurity false positives when using Fable 5.1, a significant reduction for security teams relying on AI-assisted code review and vulnerability scanning. Fable 5.1 is designed to discover software vulnerabilities but cannot be used to develop exploits for them, establishing a clear offensive boundary.
The model now includes invisible watermarking for all text it generates, with Anthropic rolling out a detection API in private preview for eligible organizations. This implements EU regulatory requirements for AI-generated content attribution and provenance tracking.
Deployment and Access Tiers
Fable 5.1 is available immediately on all platforms and carries no access restrictions. Mythos 5.1, the same model with additional safeguards, remains limited to U.S. Companies and individuals participating in Anthropic’s trusted access programs, which include Apple as part of Project Glasswing.
A new Life Sciences Verification Program grants Mythos 5.1 users access to advanced biology capabilities for scientific research, with open enrollment for scientists coming soon. Anthropic has also updated Claude Code to version 2.1.224 on macOS and Linux, allowing separate coding sessions to message each other directly without manual context copying between Terminal windows.
