nvidia2025cc-by-4.0

GraspGen: Scaling Sim2Real Grasping

A large-scale simulated grasp dataset containing over 57 million grasps computed for 8515 objects from Objaverse XL, specific to three gripper types: Franka Panda, Robotiq-2f-140, and a single-contact suction gripper.

Downloads750
Likes30

Why This Matters for Physical AI

This dataset provides a large-scale collection of simulated grasps across multiple gripper embodiments and real objects, enabling training of sim-to-real transferable grasping policies.

Technical Profile

Robot Embodiments
Franka PandaRobotiq-2f-140suction gripper
Environment
simulation
Task Types
grasping
Data Format
WebDataset
Annotation Types
grasp_success_labelsgripper_poses
License
cc-by-4.0
Part of the GraspGen family

Community Signals

Top 25% by downloads
HuggingFace Discussions1

Access

Need custom physical AI data?

Claru builds purpose-built datasets for simulation applications with dense human annotations and quality assurance.

Request a Sample Pack