---
title: univla-7b
canonical_url: "https://www.modelscope.cn/models/WangC23/univla-7b"
md_url: "https://www.modelscope.cn/models/WangC23/univla-7b.md"
repository: WangC23/univla-7b
last_updated: 2025-08-11
license: apache-2.0
pipeline_tag: robotics
tasks:
  - robotics
model_type:
  - openvla
architectures:
  - OpenVLAForActionPrediction
parameters: 7.5B
tensor_type:
  - BF16
library_name:
  - safetensors
downloads: 41
stars: 0
---

# univla-7b

> univla-7b - WangC23 在 ModelScope 开源的模型。🌏 UniVLA This is the official checkpoint of our RSS 2025 work: Learning to Act Anywhere with Task-centric Latent Actions

WangC23/univla-7b 是 ModelScope 魔搭社区上的 7.5B 参数robotics模型，采用 apache-2.0 许可。

- **Repository**: WangC23/univla-7b
- **License**: apache-2.0
- **Tasks**: robotics
- **Parameters**: 7.5B
- **Downloads**: 41
- **Stars**: 0
- **Last updated**: 2025-08-11

Source: https://www.modelscope.cn/models/WangC23/univla-7b

---

# 🌏 UniVLA
> This is the official checkpoint of our RSS 2025 work: **Learning to Act Anywhere with Task-centric Latent Actions**

#### Paper: https://arxiv.org/pdf/2505.06111
#### Code: https://github.com/OpenDriveLab/UniVLA

## 🔥 Highlights
- A recipe towards generalist policy by planning in a unified, embodiment-agnostic action space.
- A novel approach for extracting task-centric latent actions from cross-embodiment videos.
- A VLA that achieves state-of-the-art results on multiple benchmarks with compute-efficient training.

## How to use

This is the UniVLA pre-trained on our full data collection (OpenX + Ego4D). For finetuning on simulation benchmarks or your customized dataset, please visit our [official repo](https://github.com/OpenDriveLab/UniVLA).

## 📝 Citation
If you find our code or models useful in your work, please cite [our paper](https://arxiv.org/pdf/2505.06111):

```bibtex
@article{bu2025univla,
  title={Univla: Learning to act anywhere with task-centric latent actions},
  author={Bu, Qingwen and Yang, Yanting and Cai, Jisong and Gao, Shenyuan and Ren, Guanghui and Yao, Maoqing and Luo, Ping and Li, Hongyang},
  journal={arXiv preprint arXiv:2505.06111},
  year={2025}
}
```