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

GUI Navigation and Interaction on DynamicGUIBench

36.8Success Rate (Chrome)

DynamicUI (w/ Qwen3-vl-8B)

4.0412.54521.0529.555Apr 28, 2026
Updated 1mo ago

Evaluation Results

MethodLinks
36.83014.314.37.19.128.644.4408.322.1
291014.307.112.214.311.1016.715.5
2026.04
26.32214.307.17.514.30608.315.1
2026.04
21.12014.3014.39.114.304016.714.8
2026.04
18.4107.107.19.114.3040010.7
2026.04
18.433007.14.528.602016.711.7
2026.04
13.2107.107.19.1002008.1
2026.04
13.2107.107.19.128.6060010.7
2026.04
10.5014.307.16.142.902008.7
2026.04
10.510014.37.1300016.76.7
2026.04
7.9107.107.16.100208.36.7
2026.04
5.307.100000002.8