Announcing ManiSkill2: a unified, fast, and accessible benchmark - robot learning made simple! - ✅Pip installable & easily deployable - ⚡Blazingly fast visual RL support - ✨Diverse task families, objects, & 4M demonstration frames - 🖥️Interactive GUI

ManiSkill2 will also appear in #ICLR23. Moreover, the online challenge 🕹 is currently open with $20,000 in prizes 🏆! Everyone is welcome to compete at

ManiSkill2 is pip installable and directly runnable in Google Colab. It is easily configurable on any GPU workstation! Starting from a clean Python environment, you can see a robot working in 1 minute! In addition, we provide hands-on Colab tutorials

ManiSkill2 provides blazingly fast visual RL environments with our asynchronized rendering system. On a regular workstation, your robot can interact w/ the environment and collect millions of visual observations within just 10 min! *Tested w/ Intel I9-9960X & 1 NVIDIA Titan RTX

ManiSkill2 features diverse task families (single/dual arm, rigid/soft body, stationary/mobile robot), w/ 2,000+ objects & 4M demonstration frames Tasks are verified solvable You can immediately study algorithms w/o worrying about task design, asset creation, & demo collection!

Besides RGB(D) image observations, ManiSkill2 supports 3D point cloud observations natively. Try your favorite 3D learning algorithms and develop your own!

Our physics-grounded depth sensor simulator comes with realistic artifacts and is tested to have small sim2real domain gap. Train your vision models in simulator and deploy in real world with limited additional efforts!

We provide an interactive GUI to inspect and interact with the environments. See what's going on w/ your agents and what's going wrong!