Qwen-UI-Agent Technical Report: Toward Next-Generation Real-World Centric Foundation GUI Agents

Researchers introduced Qwen-UI-Agent, a GUI agent framework that combines sandbox and real-device training.

It supports both GUI and CLI actions, and is built for long-horizon, proactive workflows.

The system uses online RL with large-scale concurrent environments and achieved state-of-the-art results on several mobile benchmarks.

It also delivered competitive performance on computer-use and browser-use evaluations.