About
Windows Use revolutionizes AI agent interaction with desktops and browsers by introducing a 'big model directs, small model executes' architecture. Instead of expensive large language models consuming thousands of tokens on raw screenshots, a lightweight local model handles all the intricate clicking, typing, and verification steps on Windows and Chrome. The big model receives concise text reports and optional embedded images, saving up to 90% of context cost. Designed for simplicity, it offers a single-dependency installation, integrates with your actual Chrome browser, and supports any OpenAI-compatible model for efficient, cost-effective desktop automation.
