Reinforcement Learning & Robotics

Production-ready RL algorithms.

GPU-accelerated robot learning framework.

General-purpose physics simulator for robotics, biomechanics, and ML research. High-fidelity contact dynamics with native Python and C++ bindings. Apache 2.0 licensed.

Standard RL environment API.

End-to-end sandbox execution environment framework for agentic reinforcement learning training built on Gymnasium-compatible APIs.

Multi-agent reinforcement learning library with JAX-accelerated environments and baselines.

Reinforcement learning environments and training tools for the MicroDuck robotics platform.