Moonshot AI’s July 2026 launch of the 2.8-trillion-parameter Kimi K3 drove recent sentiment, delivering reported scores of 56% on Humanity’s Last Exam with tools and roughly 43.5–46.9% without, placing it near the top of public leaderboards but behind leading closed models such as Anthropic’s Claude Fable 5.1 and Opus 5 variants at 59–65%. The benchmark’s 2,500 expert-level questions across math, science, and humanities reward advanced reasoning and tool use, areas where Moonshot’s large context window and open-weight approach support continued fine-tuning and agentic improvements. Trader focus centers on whether additional Kimi iterations or optimizations can close the remaining gap to 50%+ or higher before year-end resolution, amid rapid frontier progress and the inherent uncertainty of model release timelines.