Alibaba’s Qwen series maintains a fast release cadence, with the current flagship Qwen3.8 Max—a 2.4-trillion-parameter sparse MoE model activating 95 billion parameters, supporting 1-million-token context and multimodal inputs—launched via API on August 3, 2026. The September 22 Apsara Conference announcement that the Qwen4 family is in training, combined with prior Max-tier updates (3.7 in May, 3.8 in August) and competitive pressure from labs such as DeepSeek and Moonshot, drives trader expectations for a 3.9+ checkpoint or early Qwen4-Max before year-end. Markets price roughly 80-90% probability by late November, reflecting Alibaba’s pattern of incremental flagship improvements focused on agentic coding and long-horizon tasks, though exact timing remains sensitive to internal training milestones and any surprise previews.