UI-Venus-2 Technical Report
The report introduces UI-Venus-2, a foundation GUI agent capable of handling mobile, web, and desktop tasks through a closed-loop reasoning-action loop. It expands environmental coverage to over 170 multilingual apps and native OSs, and uses a deep-research pipeline to generate function-grounded instructions.
For verification, the approach employs trace-level and sample-level evaluators with visual keypoints and multi-model voting to provide reliable RL signals. Safety-aware mechanisms are integrated to ensure controlled execution of consequential actions, aiming for deployable, verifiable performance.
Why it matters: By broadening environments and refining verification, UI-Venus-2 moves toward more reliable real-world GUI automation. The safety and open-source emphasis support practical deployment and community-driven improvement.
AI-assisted brief
official source
Add a comment
No account required