POOLSIDE RELEASES 118B LAGUNA S 2.1 MODEL

Laguna S 2.1 targets developers working on complex coding projects and agent-based systems that require sustained reasoning over extended sequences. The model's open-weight architecture allows developers to deploy and fine-tune it locally without proprietary restrictions. Poolside positions the release as a response to competitive pressure from Chinese open-source models including Kimi, Deepseek, and Qwen, which have gained significant traction in the AI tools market. The company joins American startups attempting to challenge this momentum, including efforts from Mira Murati's Thinking Machines and others. The 118B parameter scale represents a middle ground in the current model hierarchy, offering performance improvements over smaller models while remaining computationally more efficient than frontier-scale alternatives. Availability and pricing details for Laguna S 2.1 were not immediately specified.