poolside's latest coding agent model, laguna s 2.1, is now visible on openrouter's model board. this model features a total of 118 billion parameters, with 8 billion active parameters, and it has achieved a score of 70.2% on terminal-bench 2.1, marking a significant step in coding agent performance.
the launch of laguna s 2.1 positions poolside as a serious contender in the competitive landscape of AI coding agents. developers and operators will want to keep an eye on how this model performs in real-world applications, especially given its robust parameter count and scoring metrics.
however, with many models vying for attention, it’s crucial to remain cautious about overestimating its capabilities based on initial benchmarks alone. @OpenRouter
