Anthropic’s Claude Fable series continues to post competitive Humanity’s Last Exam results, with Fable 5.1 variants scoring 59–65% on recent October 2026 leaderboards, though Kimi K3.1 Pro currently leads at 70% in some evaluations. The benchmark’s 2,500 expert-level questions across disciplines test frontier reasoning that resists saturation, keeping scores in the 60–70% range and highlighting incremental gains from tool use and model scaling. Trader focus centers on whether the anticipated Fable 5.2+ release—forecast by prediction markets for mid-to-late November—will deliver the first clear Anthropic lead or a new high-water mark. No major Anthropic announcements have occurred in the past month, leaving competitive positioning against Moonshot, OpenAI, and Meta models as the dominant near-term variable.