Qwen3.6-35B-A3B open-sourced
AI summary
Qwen open-sourced Qwen3.6-35B-A3B, a sparse MoE model with 35 billion total parameters and only 3 billion activated parameters during inference. It significantly surpasses its predecessor on coding benchmarks such as SWE-bench and Terminal-Bench, and outperforms a 27-billion-parameter dense model; it is natively multimodal and scores 92.0 on RefCOCO.
Key Takeaways
- Qwen open-sourced Qwen3.
- 6-35B-A3B,a sparse MoE model with 35 billion total parameters and only 3 billion activated…
- It significantly surpasses its predecessor on coding benchmarks such as SWE-bench and…
- it is natively multimodal and scores 92.
- 0 on RefCOCO.
Quick Facts
- Published
- Updated
- Source
- 公众号:通义实验室(千问)
- Category
- Model Releases
- AI score
- 62
千问又推轻量 MoE 模型,35B 总参激活仅 3B,小团队能跑,但正文就一句宣传语,缺技术细节,先观望。
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: Qwen3.6-35B-A3B open-sourced. April 16, 2026. https://new.nat6.net/en/news/cmq2o71l101jnsl6n5x3b4qk1/