---
language:
- en
pipeline_tag: image-to-video
tags:
- video-generation
- image-to-video
- world-model
- robotics
- action-conditioned
- pytorch
library_name: pytorch
base_model:
- Wan-AI/Wan2.2-TI2V-5B
---

<div align="center">

<h1>Boundless-World-Model</h1>

<p align="center">
  <strong>BWM: Physically consistent, action-conditioned video world model for robotic manipulation</strong>
</p>

<p align="center">
  <a href="https://github.com/boundless-large-model/boundless-world-model"><img src="https://img.shields.io/badge/GitHub-Repository-blue?style=flat&logo=github" alt="GitHub Repository"></a>
  <a href="https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B"><img src="https://img.shields.io/badge/Base%20Model-Wan2.2--TI2V--5B-orange?style=flat&logo=huggingface" alt="Base Model"></a>
  <a href="https://huggingface.co/spaces/WorldArena/WorldArena"><img src="https://img.shields.io/badge/Benchmark-WorldArena-yellow?style=flat" alt="WorldArena"></a>
</p>

</div>

## Model Details

| Property | Value |
|----------|-------|
| **Base Model** | [Wan2.2-TI2V-5B](https://huggingface.co/Wan-AI/Wan2.2-TI2V-5B) |
| **Resolution** | 480 x 640 |
| **Frames** | 81 frames |
| **Control Signals** | Robot action trajectories |
| **Architecture** | Trainable DiT + Action Encoder |

## Usage

To use these weights, please refer to [our GitHub repository](https://github.com/boundless-large-model/boundless-world-model).

## Acknowledgements

This project builds upon the following open-source projects and benchmarks:

- Wan2.2: https://github.com/Wan-Video/Wan2.2
- DiffSynth-Studio: https://github.com/modelscope/DiffSynth-Studio
- WorldArena: https://github.com/tsinghua-fib-lab/WorldArena/
- ABot-PhysWorld: https://github.com/amap-cvlab/ABot-PhysWorld