Binyuan Hui(@huybery)
So excited to see Qwen3.5 officially out! 🎉 Congrats to everyone who made this happen. Onward and u...
5.5内容质量
TL;DR · AI 摘要
Qwen3.5系列首个开源模型Qwen3.5-397B-A17B发布，支持原生多模态，采用混合线性注意力与稀疏MoE架构。
核心要点
- Qwen3.5-397B-A17B是Qwen3.5系列首个开源权重模型
- 模型原生支持多模态，面向真实世界智能体场景训练
- 相比Qwen3-Max，解码吞吐量提升8.6至19倍
#Qwen#大模型#多模态#MoE#阿里云
打开原文Congrats to everyone who made this happen. Onward and upward!" / X
So excited to see Qwen3.5 officially out! 
Congrats to everyone who made this happen. Onward and upward!
Quote
Qwen
@Alibaba_Qwen
Feb 16
Qwen3.5-397B-A17B is here: The first open-weight model in the Qwen3.5 series. 
Native multimodal. Trained for real-world agents. 
Powered by hybrid linear attention+sparse MoE and large-scale RL environment scaling. 
8.6x–19.0x decoding throughput vs Qwen3-Max 
201
