Recent benchmark leadership by Xiaomi’s MiMo-V2.6-Pro on the Artificial Analysis Intelligence Index, combined with Alibaba’s unmatched Qwen ecosystem scale and developer adoption, underpins the tight 46.3%–33.0% implied odds between the two. Moonshot’s Kimi K3.1 Pro, the largest open-weight model released to date, sits at 10.5% as traders price in its November 4 weights drop and strong reasoning scores, while Z.ai’s GLM-5.3 and DeepSeek’s low-cost V4.1 Flash trail at single digits. The broader field remains fragmented because Chinese labs emphasize open weights, 1M-token context, and specialized strengths—cost for DeepSeek, enterprise reach for Alibaba—rather than a single frontier winner. Key near-term catalysts include Moonshot’s open release, any late-October or November benchmark updates, and the impact of proposed export rules that exempt already-released weights.