Tencent-Hunyuan/Hunyuan3D-Omni

Tencent-Hunyuan/Hunyuan3D-Omni

Description: Hunyuan3D-Omni: A Unified Framework for Controllable Generation of 3D Assets

Language: Python

License: NOASSERTION

Stars: 586

Forks: 53

Open issues: 8

Created: 2025-09-25T13:13:53Z

Pushed: 2025-10-17T10:13:05Z

Default branch: main

Fork: no

Archived: no

README:

[//]: # ( )

[//]: # ( )

[//]: # ( )

Hunyuan3D-Omni

Hunyuan3D-Omni is a unified framework for the controllable generation of 3D assets, which inherits the structure of Hunyuan3D 2.1. In contrast, Hunyuan3D-Omni constructs a unified control encoder to introduce additional control signals, including point cloud, voxel, skeleton, and bounding box.

Multi-Modal Conditional Control

• Bounding Box Control: Generate 3D models constrained by 3D bounding boxes.
• Pose Control: Create 3D human models with specific skeletal poses.
• Point Cloud Control: Generate 3D models guided by input point clouds.
• Voxel Control: Create 3D models from voxel representations.

🎁 Models Zoo

It takes 10 GB VRAM for generation.

| Model | Description | Date | Size | Huggingface | |----------------------------|-----------------------------|------------|------|-------------------------------------------------------------------------------------------| | Hunyuan3D-Omni | Image to Shape Model with multi-modal control | 2025-09-25 | 3.3B | Download |

Installation

Requirements

We test our model with Python 3.10.

pip install torch==2.5.1 torchvision==0.20.1 torchaudio==2.5.1 --index-url https://download.pytorch.org/whl/cu124
pip install -r requirements.txt

Usage

Inference

Multi-Modal Inference

python3 inference.py --control_type [--use_ema] [--flashvdm]

The control_type parameter has four available options:

point: Use point control type for inference. voxel: Use voxel control type for inference. bbox: Use bounding box control type for inference. pose: Use pose control type for inference.

The --use_ema flag enables the use of Exponential Moving Average (EMA) model for more stable inference.

The --flashvdm flag enables FlashVDM optimization for faster inference speed.

Please choose the appropriate control_type based on your requirements. For example, if you want to use the point control type, you can run:

python3 inference.py --control_type point
python3 inference.py --control_type point --use_ema
python3 inference.py --control_type point --flashvdm

Acknowledgements

We would like to thank the contributors to the TripoSG, CLAY, Trellis, DINOv2, Stable Diffusion, FLUX, diffusers, HuggingFace, CraftsMan3D, Michelangelo, Hunyuan-DiT, HunyuanVideo, HunyuanWorld-1.0, HunyuanWorld-Voyager, and PoseMaster works, for their open research and exploration.

Citation

If you use this code in your research, please cite:

@misc{hunyuan3d2025hunyuan3domni,
title={Hunyuan3D-Omni: A Unified Framework for Controllable Generation of 3D Assets},
author={Tencent Hunyuan3D Team},
year={2025},
eprint={2509.21245},
archivePrefix={arXiv},
primaryClass={cs.CV},
url={https://arxiv.org/abs/2509.21245},
}

@misc{hunyuan3d2025hunyuan3d,
title={Hunyuan3D 2.1: From Images to High-Fidelity 3D Assets with Production-Ready PBR Material},
author={Tencent Hunyuan3D Team},
year={2025},
eprint={2506.15442},
archivePrefix={arXiv},
primaryClass={cs.CV}
}

@misc{hunyuan3d22025tencent,
title={Hunyuan3D 2.0: Scaling Diffusion Models for High Resolution Textured 3D Assets Generation},
author={Tencent Hunyuan3D Team},
year={2025},
eprint={2501.12202},
archivePrefix={arXiv},
primaryClass={cs.CV}
}

@misc{yang2024hunyuan3d,
title={Hunyuan3D 1.0: A Unified Framework for Text-to-3D and Image-to-3D Generation},
author={Tencent Hunyuan3D Team},
year={2024},
eprint={2411.02293},
archivePrefix={arXiv},
primaryClass={cs.CV}
}

Star History