TexVerse: A Universe of 3D Objects with High-Resolution Textures

     
     
Yibo Zhang^1,2, Li Zhang^1,3, Rui Ma^{2 *}, Nan Cao^1,4
1Shanghai Innovation Institute ²Jilin University ³Fudan University ⁴Tongji University
* Corresponding Author

TexVerse is a large-scale 3D dataset featuring high-resolution textures. Its key characteristics include:

1. Scale & Source: TexVerse dataset has 858,669 unique 3D models curated from Sketchfab, including 158,518 with physically based rendering (PBR) materials.
2. Variants: Each model includes all high-resolution variants, resulting in a total of 1,659,097 3D instances.
3. Specialized Subsets: For the rigged and animated categories of models, we further obtain the original user-uploaded file format to prevent the loss of skeletons and animations during the format conversion of Sketchfab.
   • TexVerse-Skeleton: 69,138 rigged models.
   • TexVerse-Animation: 54,430 animated models.
4. Annotations: Detailed model annotations covering overall characteristics, structural components, and fine-grained features.

Feel free to contact me ([email protected]) or open an issue if you have any questions or suggestions.

News

• [2025-08-11] We have released the TexVerse dataset. Feel free to download it.

Download

TexVerse Dataset
Due to the file number limitations of the Hugging Face repository:

• The 2K, 4K, and 8K resolution data are hosted in TexVerse.
• The 1K resolution data are hosted separately in TexVerse-1K.

TexVerse-Animation Dataset and TexVerse-Skeleton Dataset
Jointly hosted in TexVerse-Skeleton-Animation.

License

Individual objects in TexVerse are all licensed as creative commons distributable objects. The metadata will provide the license for each object.

• CC BY: 699,075
• CC BY-NC-SA: 88,710
• CC BY-SA: 14,624
• CC BY-NC: 31,514
• CC BY-NC-ND: 11,599
• CC BY-ND: 7,008
• CC0: 6,139

BibTex

@article{zhang2025texverse,
  title     = {TexVerse: A Universe of 3D Objects with High-Resolution Textures},
  author    = {Zhang, Yibo and Zhang, Li and Ma, Rui and Cao, Nan}, 
  journal   = {arXiv preprint},
  year      = {2025}
}