Anthropic's Claude Fable 5 claims top benchmark ranking but costs double its predecessor for just a 5.7 percent performance gain.

What happened: Anthropic released Claude Fable 5, which scores 64.9 points on the Artificial Analysis Intelligence Index and holds the top position. The model costs $10 and $50 per million input and output tokens—double the $5 and $25 charged for the previous flagship Opus 4.8. A full index evaluation run costs $9,940 for Fable 5 versus $4,970 for Opus 4.8.

Why it matters: Companies must decide whether a modest 5.7 percent performance gain justifies the doubled pricing. Anthropic has followed this pattern before with Opus versions 4.8 and 4.7, raising costs steeply for small gains. The extra cost is partly driven by safety filters that reroute requests on sensitive topics back to Opus 4.8—a fallback mechanism that still counts toward billing. On Humanity's Last Exam, the fallback rate reached nine percent, pushing costs even higher than raw token pricing suggests.

What to watch: Fable 5 is available through June 22 to subscription holders (Pro, Max, Team, and Enterprise) at double the Opus rate; after that date it moves to credit-based billing, making it even pricier. The model retains the same one-million-token context window as Opus 4.8, and Anthropic says it will restore subscription access once capacity allows.