Alibaba Qwen Releases Qwen3.8-Flash: 6B Active Params, 9x Cost Reduction
千问大模型 · wechat · 2026-08-26
Alibaba Qwen released the Qwen3.8-Flash multimodal MoE model with a 125B main size but only 6B active parameters per token. It features a GDN+QSA hybrid architecture, GatedResidual multi-branch streams, and N-gram Embedding for capacity expansion. It natively supports 262k context (extendable to 1M). Training costs are reduced to 1/9 of Qwen3.7-Plus while improving coding and office tasks. Priced at 1 RMB/M input tokens and 3 RMB/M output tokens. Weights for the new architecture Qwen3.8-Flash-Next are open-sourced.
More from Models
- Reasonable post-trains Nemotron 3.5 on 3,000 Verus proofs, lifting pass@1 ~30 points to 73.3% — fhuszar · 2026-10-16
- NVIDIA's Nemotron 3.5 Super VL hits CoreWeave: 12B-active multimodal MoE at $0.25/M input tokens — wandb · 2026-10-16
- Opus 5 might be a very deep model or use looping, commentator speculates — scaling01 · 2026-10-16
- GPT-6.1 Sol audit finds 6% of arXiv papers had false main theorems, down to 1.5% recently — nitarshan · 2026-10-16
- Nvidia-backed post-training pushes Nemotron 3.5 VL past frontier on dermatology diagnosis — LoganGrasby · 2026-10-16
- Gemini 4 Argon scores 97.5% on ARC-AGI-3, second lab above 95% — patience_cave · 2026-10-16