Event ·Official first-hand

Poolside发布Laguna S 2.1，编码模型在SWE-Bench多语言领先

Poolside发布Laguna S 2.1，这是一个118B参数的MoE编码模型，每token仅激活8B参数，支持1M上下文。该模型在SWE-Bench Multilingual基准上匹配甚至超越参数规模数倍于它的模型。Laguna S 2.1采用OpenMDW-1.1开放权重许可，可在单台NVIDIA DGX Spark上运行。

Current read

Poolside出了个小参数但很强的编码模型Laguna S 2.1，8B活跃参数就在SWE-Bench多语言上吊打大模型，还开放权重，单机就能跑。

11 sources54° AI heatLast updated 2026/7/22 00:01:32

Evidence

11 sources

Primary sourcemarktechpost

2026/7/22 00:01:32View source

Related source 1NVIDIA AI

2026/7/21 17:08:07View source

Related source 2vLLM

2026/7/23 07:13:47View source

Related source 3LMSYS Org (SGLang)

2026/7/23 14:37:47View source

Related source 4elvis

2026/7/21 22:58:44View source

Related source 5Thomas Wolf

2026/7/22 08:46:49View source

Related source 6Latent Space (swyx)

2026/7/23 05:09:14View source

Related source 7Decoder

2026/7/23 12:24:53View source

Related source 8IT之家

2026/7/20 03:57:36View source

Related source 9Latent.Space

2026/7/23 19:01:32View source

Related source 10AWS Machine Learning Blog

2026/7/20 17:01:32View source

Conflict check

Dedup data only shows that these sources discuss the same event, not that they agree. There is no structured supporting or refuting evidence yet, so no conflict is inferred automatically.

Open the story