# HunyuanWorld-Voyager

Voyager generates explorable scenes as RGB-D video that follows a camera path you define, keeping the geometry consistent enough to reconstruct in 3D. It renders offline rather than responding live.

Organization: Tencent Hunyuan. Released: 2 September 2025. Kind: Explorable world.

## What you control

- A starting image
- A custom camera trajectory

## License

Weights: Available (https://huggingface.co/tencent/HunyuanWorld-Voyager). Declared license: Tencent HunyuanWorld-Voyager Community License.
Only under its own license (Tencent HunyuanWorld-Voyager Community License). Not licensed in the European Union, the United Kingdom or South Korea.

## Real-time

No real-time claim. The README reports offline generation times.

## How to run it

1. Create a Python 3.11 conda environment and install requirements.txt, flash-attn, xfuser, MoGe and utils3d.
2. Download the pretrained models listed in the README.
3. Generate camera-controlled video from an image and a camera path with the provided scripts.

## Where to try

No hosted destination verified.

## Sources

- [Project source](https://github.com/Tencent-Hunyuan/HunyuanWorld-Voyager)
- [Weights and model card](https://huggingface.co/tencent/HunyuanWorld-Voyager)

## Questions

### Is HunyuanWorld-Voyager open source?

HunyuanWorld-Voyager publishes downloadable weights under Tencent HunyuanWorld-Voyager Community License. Read the license terms before you build on it.

### Can I use HunyuanWorld-Voyager commercially?

Only under its own license (Tencent HunyuanWorld-Voyager Community License). Not licensed in the European Union, the United Kingdom or South Korea.

### Does HunyuanWorld-Voyager run in real time?

The README reports offline generation times.

### Where can I try HunyuanWorld-Voyager?

No hosted version was verified; run the published weights yourself.

### What does HunyuanWorld-Voyager take as input?

You give it a starting image and a custom camera trajectory.

### How do I run HunyuanWorld-Voyager?

Following the official instructions: Create a Python 3.11 conda environment and install requirements.txt, flash-attn, xfuser, MoGe and utils3d. Download the pretrained models listed in the README. Generate camera-controlled video from an image and a camera path with the provided scripts.

### What are alternatives to HunyuanWorld-Voyager?

Similar models in this directory include ABot-World (open weights, permissive license), AlayaWorld (open weights, custom terms) and DreamX-World (open weights, permissive license).

Checked 2026-10-23. Corrections: https://github.com/windflow-labs/world-model-directory