Anthropic launches Claude Fable 5, a publicly available version of its Mythos model, with safety guardrails that block responses in high-risk areas.

Anthropic released Claude Fable 5 on Tuesday as the first publicly available version of its Mythos model. It excels at software engineering, knowledge work, and vision, but blocks responses in high-risk areas including cybersecurity, biology, chemistry, and distillation, falling back to Claude Opus 4.8 instead.

Fable 5 is available through Anthropic's Claude API and consumption-based Enterprise plans. Through June 22, it is included at no extra cost in Pro, Max, Team, and seat-based Enterprise plans; starting June 23, it will require usage credits. Pricing is $10 per million input tokens and $50 per million output tokens.

Early data shows at least 95% of Fable sessions run entirely on the model's own responses without deferring to Opus 4.8. Third-party evaluators reported that Fable achieved a 90% on Hex's core analytics benchmark and outperformed other models on tasks including UI design and game coding.

Anthropic implemented a mandatory 30-day retention policy on all traffic to defend against attacks and jailbreaks, even for enterprises with previous zero-retention agreements. The company said over 1,000 hours of internal testing and external red-teaming produced no universal jailbreaks.