Anthropic’s Claude Fable is a version of Mythos the public can access today

Anthropic has launched Claude Fable 5, the first publicly available version of its Mythos-class AI model. While it excels at software engineering, knowledge work, and vision tasks, it comes with hard safety limits — blocking responses in high-risk areas like cybersecurity, biology, and chemistry, falling back to Claude Opus 4.8 in those cases. The model is available via the Claude API and Enterprise plans, though access on subscription plans will require usage credits after June 22. Anthropic also deployed Mythos 5 to pre-approved organizations. A notable policy change requires 30-day data retention on all traffic, even for enterprises with prior zero-retention agreements, framed as a safety measure against novel jailbreaks. Pricing is $10/million input tokens and $50/million output tokens — double Opus 4.8's price. The launch coincides with Anthropic's IPO preparations and its public warning about AI systems potentially achieving recursive self-improvement.