allenai/objaverse-xl
A dataset repository containing 10 million+ 3D objects with scripts for downloading and processing, used to train the Zero123-XL 3D foundation model.

Objaverse-XL provides over 10 million 3D objects with accompanying API scripts for downloading and processing. The dataset was used to train Zero123-XL, a foundation model for 3D synthesis that demonstrates strong zero-shot generalization. Scripts enable downloading raw assets and processing them for use in image-to-3D and text-to-3D pipelines. The project includes rendering infrastructure and shows scaling trends for 3D generation models.
Frequently asked
- What is allenai/objaverse-xl?
- A dataset repository containing 10 million+ 3D objects with scripts for downloading and processing, used to train the Zero123-XL 3D foundation model.
- Is objaverse-xl open source?
- Yes — allenai/objaverse-xl is open source, released under the Apache-2.0 license.
- What language is objaverse-xl written in?
- allenai/objaverse-xl is primarily written in Python.
- How popular is objaverse-xl?
- allenai/objaverse-xl has 1.3k stars on GitHub.
- Where can I find objaverse-xl?
- allenai/objaverse-xl is on GitHub at https://github.com/allenai/objaverse-xl.