Poolside Releases Laguna S 2.1: A Small Coding Model Outperforming 1T+ Parameter Giants

mervenoyann · x · 2026-07-22

Poolside has announced Laguna S 2.1, claiming it is the most capable agentic coding model in its weight class.

On the Terminal-Bench 2.1 benchmark, the model scored 70.2, competing directly with models 5–25x its size and outperforming several of them. On the highly challenging long-horizon DeepSWE benchmark, Laguna S 2.1 scored 40.4, beating multiple open-source models with over 1 trillion parameters. The company has also released the full trajectory of every trial in the final evaluation set on GitHub for transparency.

Related event: Poolside Releases Laguna S 2.1 Open-Weight Coding Model(10 posts)→

Original post →

More from Models

Models channel →