Xiaomi leads the market at 59% implied probability due to its late-September release of MiMo-V2.6-Pro, which tops Chinese models on the Artificial Analysis Intelligence Index at 46 while offering open weights under an MIT license and low per-task costs. This omnimodal mixture-of-experts model excels in agentic coding, real-world benchmarks like FrontierCode and Agent Arena, and multimodal tasks, building on Xiaomi’s hardware ecosystem in smartphones, robotics, and consumer devices. Rivals such as Alibaba’s Qwen3.8 Max, Z.ai’s GLM-5.3, and Moonshot’s Kimi K3 remain close on benchmarks but trail in recent momentum and open-source accessibility. With multiple labs shipping models rapidly amid compressed release cycles, traders view Xiaomi’s combination of performance, cost, and ecosystem integration as the clearest near-term edge through October.