← all releases
Poolside·Jul 21, 2026·v2.1·open source
Laguna S 2.1
Laguna S 2.1 is an advanced 118B Mixture-of-Experts (MoE) agentic coding model featuring 8B activated parameters per token. It is specifically designed for long-horizon software engineering tasks and complex reasoning, capable of operating within a 1M token context window.
AI AgentCoding AssistantOpen WeightsLLM
overview
Laguna S 2.1 is an advanced 118B Mixture-of-Experts (MoE) agentic coding model featuring 8B activated parameters per token. It is specifically designed for long-horizon software engineering tasks and complex reasoning, capable of operating within a 1M token context window.
key features
- 01Mixture-of-Experts (MoE) architecture with 118B total parameters and 8B activated parameters
- 02Supports a 1M token context window in both thinking and no-thinking modes
- 03Agentic capabilities optimized for long-horizon coding and complex software engineering
- 04High transparency with full evaluation trajectories publicly available for all benchmarks
- 05Efficient size suitable for deployment on local hardware
related productsbrowse all →
Protenixopen sourceProtenix is an open-source, trainable PyTorch implementation of AlphaFold 3, designed for high-accuracy biomolecular structure prediction. It aims to advance accessible and extensible research tools for the computational biology community. ([github.com](https://github.com/bytedance/Protenix?utm_source=openai))Z-Imageopen sourceZ-Image is an efficient image generation foundation model developed by Tongyi-MAI, designed to produce high-quality, diverse, and stylistically versatile images. It serves as a robust base for creators, researchers, and developers seeking advanced image generation capabilities.