Qwen3.8-Flash-Next

The subject name, Qwen3.8-Flash-Next, resembles the naming convention used for large language models in the Qwen family, but the supplied evidence does not mention this name, this model, or any product bearing it. The two evidence items are unrelated Chinese news pages: one reports on an OpenAI security incident involving Hugging Face, and the other discusses hair-washing frequency and scalp health research. Neither page describes a commercial brand, company, product platform, or trademark called Qwen3.8-Flash-Next, and neither provides a website, owner, headquarters, founding date, or product details for it. Because no brand-sense evidence exists, this entry cannot be written as a verified brand dossier. It is flagged for review, and no factual claims about the entity's identity, ownership, offerings, or history are asserted here.

• 2026.08.27

  Alibaba Qwen team open-sources Qwen3.8-Flash-Next multimodal MoE model[3]阿里通义千问团队开源Qwen3.8-Flash-Next多模态MoE模型chinaz.com

  Alibaba's Tongyi Qianwen team open-sourced Qwen3.8-Flash-Next, an early preview of the Qwen4 architecture. The multimodal MoE model has 125 billion total parameters with 6 billion activated per token, a 256K native context window extended to 1M via YaRN, and Day-0 support from SGLang.

Read more

1. 1OpenAI发布Hugging Face安全事件官方报告163.com
2. 2「每天洗头」到底好不好？两项研究：较高的洗发频率对亚洲人群的头皮和头发健康有益，但护发产品的使用需警惕！或会影响室内空气与健康163.com
3. 3阿里通义千问团队开源Qwen3.8-Flash-Next多模态MoE模型chinaz.com