Saved in:
Bibliographic Details
Main Authors: Yang, Yuhao, Yang, Zhen, Dou, Zi-Yi, Nguyen, Anh, You, Keen, Attia, Omar, Szot, Andrew, Feng, Michael, Ramrakhya, Ram, Toshev, Alexander, Huang, Chao, Yang, Yinfei, Gan, Zhe
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2510.17790
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918524276965376
author Yang, Yuhao
Yang, Zhen
Dou, Zi-Yi
Nguyen, Anh
You, Keen
Attia, Omar
Szot, Andrew
Feng, Michael
Ramrakhya, Ram
Toshev, Alexander
Huang, Chao
Yang, Yinfei
Gan, Zhe
author_facet Yang, Yuhao
Yang, Zhen
Dou, Zi-Yi
Nguyen, Anh
You, Keen
Attia, Omar
Szot, Andrew
Feng, Michael
Ramrakhya, Ram
Toshev, Alexander
Huang, Chao
Yang, Yinfei
Gan, Zhe
contents Computer-use agents face a fundamental limitation. They rely exclusively on primitive GUI actions (click, type, scroll), creating brittle execution chains prone to cascading failures. While API-driven agents harness rich capabilities through structured interfaces and tools, computer-use agents remain constrained to low-level visual interactions. We present UltraCUA, a foundation model that transcends this limitation through hybrid action-seamlessly unifying primitive GUI operations with high-level tool execution. Our innovation rests on four critical advances. First, an automated pipeline extracts and scales tool capabilities from software documentation and code repositories. Second, a synthetic data engine produces 17,000+ verifiable tasks capturing real-world computer-use complexity. Third, comprehensive hybrid action trajectory collection incorporates both GUI primitives and strategic tool calls. Fourth, a two-stage training methodology combines supervised fine-tuning with online reinforcement learning, enabling intelligent action selection between GUI and API. Evaluation with our 7B and 32B UltraCUA models reveals transformative performance gains. On OSWorld, UltraCUA achieves 22% relative improvement while executing 11% faster than existing approaches, averagely. Cross-domain validation on WindowsAgentArena demonstrates robust generalization with 21.7% success rate, surpassing Windows-trained baselines. The hybrid action paradigm proves essential, reducing error propagation while improving execution efficiency. This work establishes a scalable paradigm bridging primitive GUI interactions and high-level tool intelligence, enabling more resilient and adaptable computer use agents for diverse environments and complex real-world tasks.
format Preprint
id arxiv_https___arxiv_org_abs_2510_17790
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle UltraCUA: A Foundation Model for Computer Use Agents with Hybrid Action
Yang, Yuhao
Yang, Zhen
Dou, Zi-Yi
Nguyen, Anh
You, Keen
Attia, Omar
Szot, Andrew
Feng, Michael
Ramrakhya, Ram
Toshev, Alexander
Huang, Chao
Yang, Yinfei
Gan, Zhe
Computer Vision and Pattern Recognition
Computation and Language
Computer-use agents face a fundamental limitation. They rely exclusively on primitive GUI actions (click, type, scroll), creating brittle execution chains prone to cascading failures. While API-driven agents harness rich capabilities through structured interfaces and tools, computer-use agents remain constrained to low-level visual interactions. We present UltraCUA, a foundation model that transcends this limitation through hybrid action-seamlessly unifying primitive GUI operations with high-level tool execution. Our innovation rests on four critical advances. First, an automated pipeline extracts and scales tool capabilities from software documentation and code repositories. Second, a synthetic data engine produces 17,000+ verifiable tasks capturing real-world computer-use complexity. Third, comprehensive hybrid action trajectory collection incorporates both GUI primitives and strategic tool calls. Fourth, a two-stage training methodology combines supervised fine-tuning with online reinforcement learning, enabling intelligent action selection between GUI and API. Evaluation with our 7B and 32B UltraCUA models reveals transformative performance gains. On OSWorld, UltraCUA achieves 22% relative improvement while executing 11% faster than existing approaches, averagely. Cross-domain validation on WindowsAgentArena demonstrates robust generalization with 21.7% success rate, surpassing Windows-trained baselines. The hybrid action paradigm proves essential, reducing error propagation while improving execution efficiency. This work establishes a scalable paradigm bridging primitive GUI interactions and high-level tool intelligence, enabling more resilient and adaptable computer use agents for diverse environments and complex real-world tasks.
title UltraCUA: A Foundation Model for Computer Use Agents with Hybrid Action
topic Computer Vision and Pattern Recognition
Computation and Language
url https://arxiv.org/abs/2510.17790