ChainCatcher 消息，阿里巴巴 Qwen 团队将于周三发布 Qwen 3.8-Flash-Next，这是一款混合专家（MoE）模型，总参数达 1250 亿，但每个 token 仅激活 60 亿参数。团队将其定位为下一代 Qwen 4 架构的预览版本，而非最终旗舰模型。
该模型被描述为多模态，并基于即将推出的 Qwen 4 架构构建。团队表示，提前发布这一早期版本是为了让开发者能够为后续完整的 Qwen 系列做好准备。模型权重将托管在 Hugging Face 与 ModelScope 平台。
截至目前，Qwen 尚未公布该模型的基准测试成绩，也未发布与自家 Qwen 3 系列或海外竞品的对比数据，1250 亿与 60 亿参数的具体数字尚未得到官方证实。作为开放权重模型，开发者可免费下载、微调并运行，无需将数据发送至封闭 API，有助于降低托管模型的成本。
Disclaimer: Investing carries risk. This is not financial advice. The above content should not be regarded as an offer, recommendation, or solicitation on acquiring or disposing of any financial products, any associated discussions, comments, or posts by author or other users should not be considered as such either. It is solely for general information purpose only, which does not consider your own investment objectives, financial situations or needs. TTM assumes no responsibility or warranty for the accuracy and completeness of the information, investors should do their own research and may seek professional advice before investing.