Alibaba's Qwen3.8-Flash-Next releasing tomorrow with Qwen4 MoE architecture

rohanpaul_ai · x · 2026-08-25

Alibaba's Qwen team announced the upcoming release of Qwen3.8-Flash-Next tomorrow. Built on the next-gen Qwen4 architecture, the model features a multimodal Mixture-of-Experts (MoE) design with 125 billion total parameters, activating only 6 billion per token for high-efficiency performance.

Related event: Alibaba Teases Qwen3.8-Flash-Next as a Preview of Qwen4 Architecture(6 posts)→

Original post →

More from Models

Models channel →