SAN FRANCISCO (TechGenez) – Anthropic today released two new versions of its flagship Claude AI models, Claude Fable 5.1 and Claude Mythos 5.1, marking a significant step forward in artificial intelligence capabilities for coding, knowledge work, and scientific research.
The new models are essentially the same base system but are offered with different levels of safeguards. Claude Fable 5.1 is generally available, while Claude Mythos 5.1 is reserved for trusted access programs focused on cybersecurity and life sciences work.
Anthropic says the updates represent the world’s most advanced models for coding and knowledge work, with early research capabilities that offer a preview of how future AI systems may contribute to scientific discovery.
Key Improvements in Pricing and Safeguards
The company made several important improvements alongside the model upgrade:
- Pricing: Claude Fable 5.1 is estimated to cost 25% less than Fable 5 for typical workloads. For highly agentic tasks, the savings can reach up to 45%. Cached input tokens are now available at a reduced rate of 0.025 yuan per million tokens.
- Data retention: New Enterprise Frontier Safeguards (EFS) give enterprise customers complete privacy with zero data retention while still preventing adversarial misuse. Eligible customers can also use Fable 5.1 with zero data retention until EFS becomes available later this fall.
- Safeguards: False positives have been reduced, especially in cybersecurity (blocking 60% fewer false positives than before). A new access program for life sciences professionals is now open for enrollment.
Strong Performance in Benchmarks
The new models show clear gains over their predecessor. Claude Fable 5.1 sets a new standard for coding and complex problem-solving. In tests, it found the root cause of a rare crash in an investment firm’s internal systems after several years of effort by engineers and other models had failed to identify it.
The company also highlighted early scientific research capabilities, including work on safety, security, and alignment. A report published yesterday described Anthropic’s ongoing improvements in these areas.
Here, you can see how Fable 5.1 compares across various benchmarks:
Trusted Access for Specialized Work
Claude Mythos 5.1 offers more permissive safeguards for vetted users in cybersecurity and life sciences. The Cyber Verification Program (CVP) and Life Sciences Verification Program (LSVP) are expanding, with the LSVP already enrolling participants in partnership with the U.S. government.
Compliance and Future Outlook
Anthropic recently signed the EU AI Act’s Code of Practice on Transparency of AI-Generated Content, along with 190 other organisations.
As the models continue to evolve, Anthropic says they are designed to work as capable collaborators rather than complete replacements for human expertise, especially in high-stakes fields like scientific research and software development.
Conclusion
Anthropic’s launch of Claude Fable 5.1 and Claude Mythos 5.1 represents a major step forward in making AI more capable, more affordable, and more aligned with real-world needs. With improved performance in coding, research, and safety features, the new models could become increasingly valuable for developers, researchers, and enterprises as the company continues to push the boundaries of artificial intelligence.
