GUI Agent Task Completion
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
91.67Success Rate (OS)
42
Jul 1, 2026
3.1Success Rate (SR)
10
Jul 1, 2026
6.2Success Rate (SR)
10
Jul 1, 2026
18.8Success Rate (SR)
10
Jul 1, 2026
71.81VLM Eval Success Rate
4
Feb 26, 2026
42.28Success Rate (Human)
4
Feb 26, 2026