Recent model releases and benchmark results have tightened competition for third place among Chinese AI labs on arena.ai Text Arena rankings. Alibaba’s Qwen3.8-Max, a 2.4-trillion-parameter mixture-of-experts system released in August with strong open-weight ecosystem traction on Hugging Face, sits neck-and-neck with Moonshot’s Kimi K3 and the newer Kimi K3.1 Pro, which posted a 55-point Artificial Analysis Intelligence Index score and competitive coding and agentic results. DeepSeek’s ultra-low-cost V4.1 Flash updates, Xiaomi’s MiMo-V2.6-Pro benchmark leads on select open models, and Z.ai’s GLM-5.3 enterprise momentum create a fluid field where open-weight availability, developer adoption, and specialized performance metrics can shift relative positioning ahead of the late-November resolution.