Recent model releases and benchmark variants are shaping sentiment on which AI lab will post the year's highest Humanity's Last Exam (HLE) score. As of early November 2026, Kimi K3.1 Pro from Moonshot AI leads tool-augmented evaluations at 70%, while Anthropic's Claude Opus 5.5 tops Artificial Analysis HLE runs at 61.4% and Google's Gemini 4 Carbon leads the refined HLE-Diamond subset at 69.1% in high-effort configurations. These gaps reflect differences in evaluation protocols—closed-book versus tool use, text-only versus multimodal, and full set versus cleaned Diamond questions—plus competitive pressure from OpenAI's GPT-6 series and Meta's Muse models. Upcoming factors include any final 2026 releases, weight drops, or refined leaderboards before year-end resolution, with traders watching for saturation signals and independent verification on the original 2,500-question expert benchmark.
Experimental AI-generated summary referencing Polymarket data. This is not trading advice and plays no role in how this market resolves. · UpdatedView resolved

Beware of external links.
Beware of external links.
Frequently Asked Questions