AgiBot
company · ChinaScores
1 product on the map — 1 closed.
Openness
2 high confidence- license
- cc-by-nc-sa-4.0(all releases)
- access
- auto(Alpha and Beta behind an automatic Hub gate
- dataset_card
- present
Anyone can download it, but only for non-commercial use: every release is CC BY-NC-SA 4.0, and the two main releases also ask for contact details first.
- https://huggingface.co/api/datasets/agibot-world/AgiBotWorld-Beta?expand[]=downloads&expand[]=cardData&expand[]=lastModified&expand[]=gated recorded 2026-09-26
Hub API for AgiBotWorld-Beta: gated auto; the gate prompt restates the CC BY-NC-SA 4.0 terms and asks for contact details.
- https://huggingface.co/api/datasets/agibot-world/AgiBotWorld2026?expand[]=downloads&expand[]=cardData&expand[]=lastModified&expand[]=gated recorded 2026-09-26
Hub API for AgiBotWorld2026: gated false; license cc-by-nc-sa-4.0.
- https://raw.githubusercontent.com/OpenDriveLab/AgiBot-World/HEAD/README.md recorded 2026-09-26
README License and Citation: "All the data and code within this repo are under CC BY-NC-SA 4.0."
Adoption
4 medium confidenceHugging Face downloads summed over the Beta, Alpha and 2026 releases. Most of the volume is the ungated 2026 release; downloads of a dataset this size count file fetches, not training runs.
- https://huggingface.co/api/datasets?author=agibot-world&sort=downloads&limit=50&expand[]=downloads&expand[]=cardData&expand[]=lastModified&expand[]=gated recorded 2026-09-26
Hub listing for agibot-world: AgiBotWorld2026 266,412, AgiBotWorld-Beta 92,675, AgiBotWorld-Alpha 35,703 downloads in 30 days.
Capability
4 medium confidenceAgiBot World matches Open X-Embodiment's million-trajectory scale, with richer sensing, but all of it comes from AgiBot's own dual-arm robots rather than a pool of many robot bodies.
- https://raw.githubusercontent.com/OpenDriveLab/AgiBot-World/HEAD/README.md recorded 2026-09-26
README: "AgiBot World Beta: Our complete dataset featuring 1,003,672 trajectories (~43.8T)"; "1 million+ trajectories from 100 robots"; "100+ 1:1 replicated real-life scenarios across 5 target domains"; "visual tactile sensors / 6-DoF Dexterous hand / mobile dual-arm robots".