Ferret-UI Lite: Lessons from Building Small On-Device GUI Agents

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Yang, Zhen, Dou, Zi-Yi, Feng, Di, Huang, Forrest, Nguyen, Anh, You, Keen, Attia, Omar, Yang, Yuhao, Feng, Michael, Zhang, Haotian, Ramrakhya, Ram, Jia, Chao, Nichols, Jeffrey, Toshev, Alexander, Yang, Yinfei, Gan, Zhe
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916980387217408
author Yang, Zhen
Dou, Zi-Yi
Feng, Di
Huang, Forrest
Nguyen, Anh
You, Keen
Attia, Omar
Yang, Yuhao
Feng, Michael
Zhang, Haotian
Ramrakhya, Ram
Jia, Chao
Nichols, Jeffrey
Toshev, Alexander
Yang, Yinfei
Gan, Zhe
author_facet Yang, Zhen
Dou, Zi-Yi
Feng, Di
Huang, Forrest
Nguyen, Anh
You, Keen
Attia, Omar
Yang, Yuhao
Feng, Michael
Zhang, Haotian
Ramrakhya, Ram
Jia, Chao
Nichols, Jeffrey
Toshev, Alexander
Yang, Yinfei
Gan, Zhe
contents Developing autonomous agents that effectively interact with Graphic User Interfaces (GUIs) remains a challenging open problem, especially for small on-device models. In this paper, we present Ferret-UI Lite, a compact, end-to-end GUI agent that operates across diverse platforms, including mobile, web, and desktop. Utilizing techniques optimized for developing small models, we build our 3B Ferret-UI Lite agent through curating a diverse GUI data mixture from real and synthetic sources, strengthening inference-time performance through chain-of-thought reasoning and visual tool-use, and reinforcement learning with designed rewards. Ferret-UI Lite achieves competitive performance with other small-scale GUI agents. In GUI grounding, Ferret-UI Lite attains scores of $91.6\%$, $53.3\%$, and $61.2\%$ on the ScreenSpot-V2, ScreenSpot-Pro, and OSWorld-G benchmarks, respectively. For GUI navigation, Ferret-UI Lite achieves success rates of $28.0\%$ on AndroidWorld and $19.8\%$ on OSWorld. We share our methods and lessons learned from developing compact, on-device GUI agents.
format Preprint
id arxiv_https___arxiv_org_abs_2509_26539
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Ferret-UI Lite: Lessons from Building Small On-Device GUI Agents
Yang, Zhen
Dou, Zi-Yi
Feng, Di
Huang, Forrest
Nguyen, Anh
You, Keen
Attia, Omar
Yang, Yuhao
Feng, Michael
Zhang, Haotian
Ramrakhya, Ram
Jia, Chao
Nichols, Jeffrey
Toshev, Alexander
Yang, Yinfei
Gan, Zhe
Computer Vision and Pattern Recognition
Computation and Language
Machine Learning
Developing autonomous agents that effectively interact with Graphic User Interfaces (GUIs) remains a challenging open problem, especially for small on-device models. In this paper, we present Ferret-UI Lite, a compact, end-to-end GUI agent that operates across diverse platforms, including mobile, web, and desktop. Utilizing techniques optimized for developing small models, we build our 3B Ferret-UI Lite agent through curating a diverse GUI data mixture from real and synthetic sources, strengthening inference-time performance through chain-of-thought reasoning and visual tool-use, and reinforcement learning with designed rewards. Ferret-UI Lite achieves competitive performance with other small-scale GUI agents. In GUI grounding, Ferret-UI Lite attains scores of $91.6\%$, $53.3\%$, and $61.2\%$ on the ScreenSpot-V2, ScreenSpot-Pro, and OSWorld-G benchmarks, respectively. For GUI navigation, Ferret-UI Lite achieves success rates of $28.0\%$ on AndroidWorld and $19.8\%$ on OSWorld. We share our methods and lessons learned from developing compact, on-device GUI agents.
title Ferret-UI Lite: Lessons from Building Small On-Device GUI Agents
topic Computer Vision and Pattern Recognition
Computation and Language
Machine Learning
url https://arxiv.org/abs/2509.26539