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

Web Navigation on WebArena Lite v2

28.6Average Success Rate

GPT-4o + ScaleCUA-7B

-0.3127.19414.722.206Feb 25, 2026
Updated 7d ago

Evaluation Results

MethodLinks
2026.02
28.6-----
2026.02
26.631.717.335.526.125
2026.02
24.429.210.634.230.117.9
2026.02
2434.210.626.316.533.6
2026.02
23.928.315.427.618.830.7
2026.02
23.4-----
2026.02
23.2-----
2026.02
22.62510.626.326.122.9
2026.02
22.6-----
2026.02
20.8-----
2026.02
17.221.77.713.216.523.6
2026.02
16.725.89.618.417.612.1
2026.02
15.6-----
2026.02
15.3155.817.11719.3
2026.02
11.917.55.810.513.110.7
2026.02
11.7-----
2026.02
5.8-----
2026.02
4.98.312.67.42.9
2026.02
0.81.7001.70