Description
Laguna by Poolside is a cutting-edge AI platform designed to advance artificial general intelligence (AGI) through innovative foundation models. Based in San Francisco, Poolside was founded in April 2023 with the mission to drive abundance for humanity by creating AGI. Laguna models are optimized for quality, speed, and efficiency, offering frontier-class reasoning capabilities at a mid-size cost. The Laguna s 2.1 model features 118 billion parameters, while the Laguna xs 2.1 model is designed for agentic coding tasks with 33 billion parameters.
Read More
Poolside emphasizes the importance of reinforcement learning (RL) as a key scaling axis for model intelligence. RL allows models to learn from new experiences and reverse engineer human knowledge, overcoming the finite nature of web data. The platform orchestrates millions of coding environments, providing agents with a safe playground to refactor, debug, and extend code while receiving immediate feedback from compilers and test suites.
Laguna models are part of Poolside's broader vision to achieve AGI by decompressing human experiential learning and introducing agent experiential learning. The platform leverages RL-driven exploration to provide diverse trajectories from real-world interactions, enhancing model capabilities through scale and compute. Poolside's approach integrates language, vision, spatial understanding, and physical interaction, aiming to extend competence across modalities.
The platform is open for research preview, offering insights into the engineering and ideas behind its models. Poolside's commitment to open-weight foundation models and reinforcement learning positions Laguna as a significant player in the pursuit of AGI.
- CategoryAI Coding Assistants
- DeploymentCloud
- PlatformsWeb
Laguna by Poolside's Core Features
Frontier-class reasoning
Agentic coding model
Open-weight foundation models
Reinforcement learning integration
Efficient parameter optimization
Real-world interaction data
Multimodal learning capabilities
Research preview availability
How to use Laguna by Poolside?
Configure: Set up the model parameters
Use: Deploy models for reasoning tasks
Optimize: Enhance performance with RL feedback
Expand: Integrate multimodal capabilities
Laguna by Poolside's Use Cases
- Reasoning Tasks
- Coding Optimization
- Multimodal Learning
- Research Preview
- AGI Development
