OpenAI’s GPT-6 Astra and related variants currently trail Anthropic’s Claude Opus 5.5 and Fable 5.1 on Humanity’s Last Exam (HLE) leaderboards, posting roughly 54.7–57.2% accuracy versus Claude’s 61–67% across text-only, tool-augmented, and verified subsets as of mid-October 2026. The 2,500-question benchmark, created by the Center for AI Safety and Scale AI, tests expert-level reasoning across disciplines and remains unsaturated, allowing rapid gains from new model releases, higher reasoning effort, or tool integration. Trader sentiment reflects OpenAI’s competitive release cadence and historical catch-up ability before year-end, tempered by Anthropic’s current lead and the short window to December 31. Key upcoming catalysts include any GPT-6 or Astra updates and official HLE leaderboard revisions.