Qwen4

Timeline

1. 2026-11-13T19:07:08.264Z · Evidence links · GitHub search · Alibaba · pulls · vllm-project/vllm-torchtpu · @DRZDONG
2. 2026-11-13T19:07:08.264Z · Evidence links · GitHub search · Alibaba · pulls · avifenesh/memra · @avifenesh
3. 2026-11-13T19:07:08.264Z · Evidence links · GitHub search · Alibaba · pulls · NVIDIA/TensorRT-LLM · @kaiyux
4. 2026-11-13T17:22:19.091Z · Evidence links · Qwen · official article bodies
5. 2026-11-09T12:09:34.176Z · Catalogue listing · GitHub · QwenLM/qwen-code · code

   This code reference does not establish a working model identifier.

6. 2026-11-05T17:57:47.304Z · Evidence links · GitHub · QwenLM/qwen-code · PR
   Excerpt

   #15762 fix(core): size the qwen4 generation at 1M

7. 2026-08-26T11:23:48.753Z · Alex's posts · Alex Getman
   Excerpt

   🚨 QWEN3.8 FLASH NEXT DROPS TODAY A ModelScope description that was quickly taken down reportedly revealed: • 125B main parameters • 51B N gram embeddings • Only 6B active per token • Multimodal MoE architecture The model is also an official preview of the upcoming Qwen4 architecture.