Anthropic has launched Claude Fable 5.1 and Mythos 5.1, two advanced AI models designed to improve coding performance and knowledge work while offering significant cost reductions for users.

Key Points

• Claude Fable 5.1 outperforms previous versions and OpenAI’s GPT-5.6 Sol across multiple industry benchmarks.
• Mythos 5.1 is restricted to U.S.-based participants in trusted access programs, including a new Life Sciences Verification Program.
• Fable 5.1 reduces cybersecurity false positives by 60 percent and includes new invisible watermarking for generated text.
• Pricing for Fable 5.1 is 25 to 45 percent lower than the previous version due to reduced cache-read costs.
• Anthropic is launching a private preview of a detection API to identify AI-generated content in compliance with EU regulations.

Why it Matters

These model updates signal a shift toward more efficient, specialized AI tools that prioritize both high-level scientific research and enterprise-grade security. By lowering costs and improving accuracy, Anthropic is positioning its technology to become a more competitive standard for complex software development and data-intensive professional tasks.