Isaac Lab
GPU-accelerated robot learning framework on Isaac Sim.
Open-source framework for training robot policies with reinforcement and imitation learning in massively parallel, physically accurate simulation built on NVIDIA Isaac Sim.
- Vendor
- NVIDIA
- Category
- Robotics & physical AI
- Pricing
- Open source
- Open source
- Yes
- License
- BSD-3-Clause
- Platforms
- Linux, Windows, Python
- Website
- developer.nvidia.com/isaac/lab
Features
- Parallel GPU simulation
- RL & imitation learning
- Sim-to-real workflows
Best for
- Agents
- Research
Works with
- Cosmos 3 — NVIDIA. NVIDIA’s open omni-modal world foundation model for physical AI.
- Isaac GR00T N1.7 — NVIDIA. Open 3B cross-embodiment VLA for humanoids and manipulators.
More robotics & physical ai
- MuJoCo — Google DeepMind. Fast, accurate physics engine for robotics research.
- LeRobot — Hugging Face. Open-source models, datasets and tools for real-world robotics.
- openpi — Physical Intelligence. Open-source π0 / π0.5 robot foundation models and training code.
- Genesis — Genesis Embodied AI. Generative physics platform for robotics and embodied AI.