benchmark · first seen 8 Oct 2026
OSWorld
Also: OSWorld benchmark
A benchmark for evaluating multimodal agents on open-ended computer tasks across real desktop applications and operating systems.
Mentions per week
PapersCodeDiscussion
Earliest sources
- SoftwareGym: Scaling Computer-Use Agent Training with Diverse Synthesized Software EnvironmentsPapers · 8 Oct 2026 · Tian Jin, Wenhao Wang, Jing Kang, Haoting Shi, Rui Ye, Yijun Lu, Siheng Chen
- CUA-Lite: An Open Platform for Computer Use AgentsPapers · 19 Oct 2026 · Zhanhui Zhou, Weichen Zhang, Haoran Liu, Lingjie Chen, Tianneng Shi, Kevin Lin,…
- Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved HarnessesPapers · 29 Oct 2026 · Xiaoyi Zhang, Haoqing Wang, Xinjie Zhang, Xiang Ming, Xing Wang, Senqiao Yang, K…
Latest sources
- Mage-VL-Cowork: A Full Training Recipe to Learn Hybrid Computer Use From Evolved HarnessesPapers · 29 Oct 2026 · Xiaoyi Zhang, Haoqing Wang, Xinjie Zhang, Xiang Ming, Xing Wang, Senqiao Yang, K…
- CUA-Lite: An Open Platform for Computer Use AgentsPapers · 19 Oct 2026 · Zhanhui Zhou, Weichen Zhang, Haoran Liu, Lingjie Chen, Tianneng Shi, Kevin Lin,…
- SoftwareGym: Scaling Computer-Use Agent Training with Diverse Synthesized Software EnvironmentsPapers · 8 Oct 2026 · Tian Jin, Wenhao Wang, Jing Kang, Haoting Shi, Rui Ye, Yijun Lu, Siheng Chen