Anthropic’s Claude Fable 5 brings massive coding power to the public

Anthropic’s launch of Claude Fable 5 marks one of the most consequential AI releases of the year – not simply because of the model’s capabilities, but because it formalizes a new frontier-model access regime. Fable 5 is Anthropic’s first generally available Mythos-class model, while Mythos 5 remains restricted-access. 

Anthropic says both are the same underlying system, with Fable wrapped in additional safety controls. The model reportedly delivers a major capability jump over Opus 4.8, including a leap on FrontierCode Diamond from 13.4% to 29.3%, while retaining a 1 million-token context window. 

Early hands-on tests prove the model’s raw capability. Wharton professor Ethan Mollick noted Fable 5 can run autonomous workflows for over 9 hours on a single design document, and Stripe used it to pull off a 50-million-line Ruby migration in one day. Simon Willison called it “slow, expensive and capable.” The API costs mirror this scale at $10 per million input and $50 per million output tokens. 

Because running these heavy reasoning loops is so expensive, Anthropic is pulling Fable 5 from flat-rate consumer subscriptions on June 22, moving it behind a strict credit-gated paywall. Furthermore, Mythos-class models dismantle standard corporate data boundaries by entirely eliminating Zero Data Retention (ZDR) options in favor of a mandatory 30-day logging policy. 

Anthropic projects this hidden throttling impacts just 0.03% of traffic, but the open-source community has strongly condemned it as anti-competitive ladder-pulling. Critics argued that secretly downgrading performance injects invisible bugs into research code and destroys any trust you have in a paid API. On the flip side, supporters argue that putting the brakes on automated self-improvement loops is just common business sense when AI capabilities are accelerating this fast