Alibaba's Qwen releases Qwen3.8-Flash-Next, previewing Qwen4 architecture

Qwen3.8-Flash-Next is a 125B multimodal MoE with 51B N-gram embeddings and only 6B active parameters per token. It reportedly beats Claude Opus 4.6 Max on 8 of 9 benchmarks. Weights and an FP8 version are scheduled for open-source release on Aug. 26.