Open-Source Robotics Learning Datasets

A curated catalog of open-source datasets for robot manipulation, imitation learning, and reinforcement learning — with links to official sources.

Catalog

Datasets for Robot Learning

Each dataset has a dedicated page with description, scale, access links, and citations.

76K trajectories, 350 hours, 86 tasks. In-the-wild manipulation from 50 collectors across 564 scenes. TensorFlow Datasets, Hugging Face.

60K trajectories, 24 environments, 13 manipulation skills. Low-cost WidowX robot. Natural language labels, multi-task learning.

1M+ episodes, 22 robot types, 500+ skills. Unified RLDS format. RT-X models. 33 institutions.

Bimanual teleoperation. ALOHA-Cosmos-Policy, baseline datasets. HDF5, Hugging Face. Open hardware.

130 tasks, 65K demos. Lifelong learning benchmark. Spatial, object, goal suites. RoboSuite simulation.

15M frames, 7 robot platforms. Multi-robot transfer. Sawyer, Franka, Baxter, Fetch, WidowX.

Framework + datasets. MimicGen: 50K demos from 200 human demos. Simulation + real. MIT license.

Standardized format + hub. DROID-100, ALOHA, SO-100. PyTorch, streaming. "ImageNet of robotics."

We collect high-quality, learning-ready data for your specific tasks and hardware.

Request Data Contact Us