Chinese AI labs continue releasing frequent updates, with Moonshot AI’s Kimi K3.1 Pro, DeepSeek V4.1 Flash, Alibaba’s Qwen3.8 Max, and Z.ai’s GLM-5.3 all scoring in the 44–46 range on the Artificial Analysis Intelligence Index as of late October 2026. These trail leading closed U.S. models such as Claude Opus 5.5 (58) by a consistent margin, though Chinese open-weight models now dominate that category and drive significant developer adoption. Alibaba recently confirmed Qwen 4 is in training with larger successors planned, while proposed export controls on advanced models exempt already-released weights. Traders are weighing whether any Chinese company can close the frontier gap before year-end amid rapid iteration but persistent benchmark shortfalls on private evaluations.