Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

GUI Agent Interaction research benchmarks · SOTA2 Research

GUI Agent Interaction

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

42.5Average Accuracy

27

Jun 15, 2026

Memory-World (test)

82.6T-Acc

22

May 29, 2026

Claude 4.5 Sonnet

42.9Success Rate (Max Steps: 15)

16

Jun 24, 2026

MobileWorld GUI-Only

Gemini-3-Pro + UI-Ins-7B

55.6SR

14

Apr 14, 2026

OSWorld w/o Loop

GPT-4o + ScaleCUA-7B

29.5AUV

5

Feb 26, 2026

OSWorld w/ Loop

Claude3.7-Sonnet

6.9AUV

5

Feb 26, 2026

Popular datasets

OSWorld MobileWorld Memory-World

Follow for updates

Join SOTA2 Contact