PulseAugur
EN
LIVE 10:04:02

New 4B hybrid agent Mage-VL-Cowork achieves SOTA on OSWorld-Verified benchmark

By PulseAugur Editorial · [1 sources] ·

Researchers have developed Mage-VL-Cowork, a 4B parameter hybrid computer-use agent (CUA) designed to seamlessly integrate visual GUI interaction with code and terminal execution. The training process involved a five-stage curriculum, utilizing Mage-CUA-Forge to synthesize over 293,000 desktop and web tasks across 6,915 skills, significantly reducing per-task expenditure. This approach led to Mage-VL-Cowork-4B achieving state-of-the-art performance for its parameter size on the OSWorld-Verified benchmark, demonstrating efficient reasoning and advancing on-device intelligence. AI

IMPACT Advances on-device intelligence and efficient reasoning capabilities for hybrid computer-use agents.

RANK_REASON The cluster contains a research paper detailing a new AI model and its training methodology. [lever_c_demoted from research: ic=1 ai=1.0]

Read on arXiv cs.CL →

AI-generated summary · Google Gemini · from 1 sources. How we write summaries →

New 4B hybrid agent Mage-VL-Cowork achieves SOTA on OSWorld-Verified benchmark

How we ranked this

Signal score
1 / 100
Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket
Tool
The cluster contains a research paper detailing a new AI model and its training methodology. [lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration
Single-source cluster
Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics
paper, model release
Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance
High
Clearly on-topic for AI-industry coverage.
Story freshness
1 days old
Coverage has settled into its steady-state source set.

Full methodology in our editorial standards.

COVERAGE [1]

  1. arXiv cs.CL TIER_1 English(EN) · Xiaoyi Zhang, Haoqing Wang, Xinjie Zhang, Xiang Ming, Xing Wang, Senqiao Yang, Kaichen Zhang, Zhaoyang Jia, Andong Tan, Xinjiang Wang, Jinghao Guo, Yifei Shen, Xiao Li, Peng Zhang, Xun Guo, Jiahao Li, Shicheng Zheng, Jinglu Wang, Zongyu Guo, Wenxuan Xie,… ·

    Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved Harnesses

    arXiv:2610.40747v1 Announce Type: new Abstract: Effective computer-use agents (CUAs) should seamlessly combine visual GUI interaction with code and terminal execution within a single policy. We present Mage-VL-Cowork, a 4B hybrid CUA trained from a general multimodally aligned ch…

RELATED ENTITIES

RELATED TOPICS