Qwen(@Alibaba_Qwen)2026年4月22日

LM Performance：With only 27B parameters, Qwen3.6-27B outperforms the Qwen3.5-397B-A17B (397B total /...

8.0Score

用这条生成生成视频方案

LM Performance：With only 27B parameters, Qwen3.6-27B outperforms the Qwen3.5-397B-A17B (397B total /...

AI 深度提炼

在 SWE-bench 和 Terminal-Bench 等基准中表现优异。
性能优于参数量大 15 倍的 Qwen3.5-397B-A17B。
适合高要求的代码生成与分析任务。

#Qwen#编码模型#AI性能

打开原文

Qwen on X: "LM Performance：With only 27B parameters, Qwen3.6-27B outperforms the Qwen3.5-397B-A17B (397B total / 17B active, ~15x larger!) on every major coding benchmark — including SWE-bench Verified (77.2 vs. 76.2), SWE-bench Pro (53.5 vs. 50.9), Terminal-Bench 2.0 (59.3 vs. 52.5), and https://t.co/kPJKh8ablz" / X

Don’t miss what’s happening

People on X are the first to know.

Post

See new posts

Conversation

![Image 1: Square profile picture](http://x.com/Alibaba_Qwen)

Qwen

@Alibaba_Qwen

LM Performance：With only 27B parameters, Qwen3.6-27B outperforms the Qwen3.5-397B-A17B (397B total / 17B active, ~15x larger!) on every major coding benchmark — including SWE-bench Verified (77.2 vs. 76.2), SWE-bench Pro (53.5 vs. 50.9), Terminal-Bench 2.0 (59.3 vs. 52.5), and SkillsBench (48.2 vs. 30.0). It also surpasses all peer-scale dense models by a wide margin.

![Image 2: Image](http://x.com/Alibaba_Qwen/status/2046939775924584577/photo/1)

1:10 PM · Apr 22, 2026

42.2K Views

389