Anthropic Unveils Claude Fable 5.1 and Claude Mythos 5.1: Advanced AI Models for Coding and Scientific Research Anthropic has launched two new iterations of its AI models, Claude Fable 5.1 and Claude Mythos 5.1, positioning them as the most advanced tools for coding, knowledge work, and scientific research. The models are functionally identical but differ in their safeguard configurations, with Fable 5.1 available to the general public and Mythos 5.1 reserved for trusted access programs focused on high-risk domains like cybersecurity and life sciences. The updates reflect Anthropic’s efforts to address user feedback on cost, data privacy, and safety measures. Fable 5.1 introduces significant cost reductions, with pricing estimated to be 25% lower than its predecessor for typical workloads, particularly when billed by token usage. This reduction stems from optimizations in cache read efficiency, which can lead to savings of up to 45% for highly agentic tasks. The model also features Enterprise Frontier Safeguards (EFS), a new system allowing enterprise customers to store data in cloud infrastructure they fully control, ensuring zero data retention while maintaining robust security. EFS will roll out in phases starting later this fall, with eligible customers able to use Fable 5.1 with zero data retention until the system is available. Safeguard improvements focus on reducing false positives, particularly in cybersecurity. Fable 5.1’s new safeguards block 60% fewer false positives than previous versions, partly due to its ability to identify software vulnerabilities without developing exploits. In biology, Anthropic has partnered with the U.S. government to create an access program for Mythos 5.1’s advanced capabilities, with enrollment expected to open soon for scientists.#nasa #anthropic #claude_fable_51 #claude_mythos_51 #millennium_investment_firm
