Traders see a tightly contested field for third place on the Text Arena math leaderboard by late October, with Alibaba holding the highest implied probability at 29.5% amid strong recent performance from its Qwen series on competition-level math benchmarks. Google follows at 21% thanks to ongoing refinements to Gemini models and specialized systems like AlphaProof, while OpenAI at 15.5% benefits from steady large language model scaling but trails specialized math efforts. Moonshot and Meta round out top contenders as Chinese labs leverage targeted training on mathematical reasoning to close gaps with Western frontier models. The even spread across additional labs reflects rapid iteration cycles, benchmark volatility, and the absence of decisive new releases in the past month, leaving room for capability jumps or slips before resolution.