Alibaba Ships Qwen3.8-Flash-Next, an Open-Weight Preview of the Qwen4 Architecture

Alibaba's Qwen team open-sourced Qwen3.8-Flash-Next, a 125B-parameter multimodal MoE model that previews Qwen4's architecture with hybrid attention and N-gram embeddings, activating just 6B parameters per token.