Xiaomi's late-September launch of its MiMo-V2.6-Pro open-weight model, which quickly topped independent benchmarks after low-cost training, has edged trader consensus toward the company at 55% implied probability. This release builds on Xiaomi's hardware integration strengths and follows earlier waves from Moonshot's Kimi series, DeepSeek, and Alibaba's Qwen updates, where cost-efficient large language models and agent capabilities have narrowed gaps with Western labs. With many listed contenders clustered near 50% and smaller players like Z.ai or MiniMax below 10%, the market reflects uncertainty over evaluation criteria such as benchmark scores, developer adoption, or multimodal performance by month-end. No major regulatory or partnership shifts have emerged in the past two weeks to alter positioning.