Windows-Use: An LLM Agent for Windows GUI Automation via Native APIs
tom_doerr · x · 2026-08-11
The GitHub trending project Windows-Use is an AI agent that controls the Windows OS at the GUI layer. Instead of relying on computer vision models, it reads the Windows UI Automation API and uses LLMs to decide actions.
Key Capabilities:
- Supports opening, switching, and resizing app windows, as well as clicking, typing, scrolling, dragging, and using shortcuts.
- Capable of running PowerShell commands, reading outputs, and scraping web pages via the browser accessibility tree.
- Features persistent memory across steps and voice input/output (STT/TTS).
More from coding & agent
- Spotify Open-Sources Xirp, Sparking Debate on AI Cost Reduction vs Product Strategy — gdequeiroz · 2026-08-11
- Cloudflare's New CI Experiment Supports Importing GitHub Actions — ritakozlov · 2026-08-11
- Swarms Weekly: Auto-Generate Multi-Agent Teams from Single Tasks, MCP Portal Launch — KyeGomezB · 2026-08-11
- DeepMind's Scaling Laws for Multi-Agent Systems: More Agents Can Degrade Performance — KyeGomezB · 2026-08-11
- Hermes Agent: Open-Source Framework for Automated AI Agent Skill Management — KhuyenTran16 · 2026-08-11
- Enterprise AI Bottleneck: Companies Rush to Use AI Before Building a Harness — jacob_posel · 2026-08-11