Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

UI Agent Evaluation on OSWorld

40.3SR (15 Steps)

Qwen-3-VL-32B

0.05210.50120.9531.399May 11, 2026
Updated 22d ago

Evaluation Results

MethodLinks
2026.05
40.39.142.818.942.527.315,229
2026.05
39.89.342.719.642.427.811,346
2026.05
35.89.340.422.239.229.615,198
2026.05
35.29.539.822.638.63011,318
2026.05
33.210.441.121.44128.44,133
2026.05
32.18.335.919.83426.46,963
2026.05
30.911.635.619.235.325.815,171
2026.05
30.81034.423.632.530.915,075
2026.05
30.79.734.522.732.329.715,071
2026.05
30.69.834.122.932.229.811,078
2026.05
30.510.134.123.832.231.111,193
2026.05
30.211.935.119.834.826.211,284
2026.05
29.812.834.220.733.927.34,061
2026.05
28.68.9322130.527.95,074
2026.05
27.910.632.423.63130.84,117
2026.05
2610.729.324.328.828.215,141
2026.05
25.510.928.524.628.528.611,227
2026.05
24.111.427.325.227.129.14,058
2026.05
23.811.227.424.62631.83,994
2026.05
21.111.424.924.624.631.415,032
2026.05
20.911.724.42524.232.611,159
2026.05
20.712.223.725.623.532.93,967
2026.05
9.511.610.225.610.933.115,136
2026.05
9.211.81025.910.733.511,221
2026.05
8.512.69.726.510.334.14,094
2026.05
4.112.44.9275.53515,183
2026.05
3.912.64.727.35.335.411,307
2026.05
3.613.44.428.1536.24,109
2026.05
312.53.8274.435.215,087
2026.05
2.912.73.627.44.235.611,238
2026.05
2.513.83.328.53.936.74,067
2026.05
2.313.22.728.13.536.515,062
2026.05
1.913.62.528.43.237.111,176
2026.05
1.614.52.229.62.738.24,013