Welcome to OSWorld!

2025-07-28: Upgrade! Introducing OSWorld-Verified with comprehensive improvements: fixed community-reported examples, AWS support reducing evaluation time to within 1 hour, and updated benchmark results. Please compare your OSWorld results with the new benchmark results when running the latest version.

Website | Paper | Code | Data | Data Viewer | Slides | Twitter | Discord

• Installation
  • Overview
  • Install Provider
  • Quick Start
  • Proxy Guideline
  • Google Account Guideline
• User Guides
  • Overview
  • DesktopEnv Interface Documentation
  • OSWorld Task Examples Explanation
  • Adding New Agents to OSWorld
  • Adding New Tasks to OSWorld
  • Run OSWorld
  • Public Evaluation Platform User Guide
  • Server setup
• Community
  • FAQ
  • To Contribute
  • OSWorld Authors
  • Vision and Roadmap
  • Projects using OSWorld
• Reference
  • Reference Paper