List your company

Web Navigation on WebArena Lite v2

28.6Average Success RateGPT-4o + ScaleCUA-7B
-0.3127.19414.722.206Feb 25, 2026

Evaluation results

MethodLinks
GPT-4o + ScaleCUA-7BModel Type=Agent Framework, Planner=GPT-4o, Steps=152026.0228.6—————
GUI-Libra-8BModel Type=Native, Steps=152026.0226.631.717.335.526.125
GUI-Libra-4BModel Type=Native, Steps=152026.0224.429.210.634.230.117.9
ScaleCUA-32BModel Type=Native, Steps=152026.022434.210.626.316.533.6
ScaleCUA-7BModel Type=Native, Steps=152026.0223.928.315.427.618.830.7
UI-TARS-72B-DPOModel Type=Native, Steps=152026.0223.4—————
GPT-4o + UGround-V1-7BModel Type=Agent Framework, Planner=GPT-4o, Steps=152026.0223.2—————
GUI-Libra-7BModel Type=Native, Steps=152026.0222.62510.626.326.122.9
GPT-4o + UI-TARS-1.5-7BModel Type=Agent Framework, Planner=GPT-4o, Steps=152026.0222.6—————
UI-TARS-1.5-7BModel Type=Native, Steps=152026.0220.8—————
ScaleCUA-3BModel Type=Native, Steps=152026.0217.221.77.713.216.523.6
GUI-Libra-3BModel Type=Native, Steps=152026.0216.725.89.618.417.612.1
Qwen2.5-VL-72BModel Type=Native, Steps=152026.0215.6—————
Qwen3-VL-8BModel Type=Native, Steps=152026.0215.3155.817.11719.3
Qwen3-VL-4BModel Type=Native, Steps=152026.0211.917.55.810.513.110.7
InternVL3.5-241B-A28BModel Type=Native, Steps=152026.0211.7—————
Aguvis-72BModel Type=Native, Steps=152026.025.8—————
Qwen2.5-VL-7BModel Type=Native, Steps=152026.024.98.312.67.42.9
Qwen2.5-VL-3BModel Type=Native, Steps=152026.020.81.7001.70