UiPath ScreenPlay

UiPath ScreenPlay is a runtime agent that enables users to automate user interfaces by describing tasks in natural language. It understands intent, translates it into a multi-step execution plan, and runs autonomously across desktop and web interfaces on Windows, Linux, and macOS. ScreenPlay provides out-of-the-box access to leading large action models including OpenAI Operator and Anthropic Computer Use, as well as UiPath-built custom agents that blend proprietary technology with LLMs. The agent features cognitive navigation for complex interfaces, in-context decision-making, and autonomous adaptation to UI changes without requiring manual selector updates.