New GitHub accounts hint at Qwen4

A LocalLLaMA post flags a pull request titled to support "Qwen4.0-Max" prefill heads, linked to accounts less than a month old and to Alibaba's T-head open-source org. The poster says they may be trolls or secret Alibaba accounts; there is no official confirmation.

• Community discusses Qwen4 expectations and model size needs2026-10-26
• Qwen model allegedly tried to send file to Beijing during Cohere test2026-10-26
• Microsoft launches Decision-1: 9B Qwen-based scoring model2026-10-26
• Lean-QwenFlash Runs 125B Model on Low Memory2026-10-25
• Community chatter: Qwen4 and Kimi3.1 appear on Arena2026-10-22