Qwen3.8-Flash-Next open-sourced: an early preview of the Qwen4 architecture
Original title:Qwen3.8-Flash-Next: A New Architecture, Towards Ultimate Cost-Efficiency
AI summary
Qwen open-sourced Qwen3.8-Flash-Next, a multimodal MoE model and an early preview of the Qwen4 architecture. The model adopts four upgrades including GDN + QSA hybrid attention, has 125B total parameters with 6B activated per token, costs about 1/9 of Qwen3.7-Plus to train, and is stronger at coding and office tasks.
Key Takeaways
- Qwen open-sourced Qwen3.
- 8-Flash-Next,a multimodal MoE model and an early preview of the Qwen4 architecture.
- The model adopts four upgrades including GDN + QSA hybrid attention,has 125B total parameters with…
- 7-Plus to train,and is stronger at coding and office tasks.
Quick Facts
- Published
- Updated
- Source
- Qwen:Blog Retrieval(API)
- Category
- Model Releases
- AI score
- 83
这次提前开放 Qwen4 架构权重,关键在于 QSA 稀疏注意力与 N-gram 查表参数同时降低长上下文成本、扩大容量,为评估下一代模型提供早期样本。
Related entities
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: Qwen3.8-Flash-Next open-sourced: an early preview of the Qwen4 architecture. August 26, 2026. https://new.nat6.net/en/news/cmta2veap03nmrolwxllvp4ay/