List your company

Web Agent Interaction on RUC suite (test)

60Success Rate (News)w/ DUDE
10.0823.043648.96May 10, 2026

Evaluation results

MethodLinks
w/ DUDEAgent Base Model=UI-TARS-1.5-7B, Evaluator Model=UI-TARS2026.056003.465824.445423.626022.96581.53.02
w/ DUDEAgent Base Model=GLM-4.6V-Flash, Evaluator Model=UI-TARS2026.056003.566824.224205.177243.1660.51.54.02
w/ DUDEAgent Base Model=Qwen3-VL-4B, Evaluator Model=UI-TARS2026.055622.887403.56404.426024.3863.50.53.85
VanillaAgent Base Model=UI-TARS-1.5-7B2026.05421618.16363217.54402416.56562211.9643.523.516.06
w/ DUDEAgent Base Model=Qwen3-VL-4B, Evaluator Model=Qwen-2B2026.052006.084404.963406.363606.0433.505.86
w/ DUDEAgent Base Model=UI-TARS-1.5-7B, Evaluator Model=Qwen-2B2026.052004.825004.043803.783404.0635.504.18
w/ DUDEAgent Base Model=GLM-4.6V-Flash, Evaluator Model=Qwen-2B2026.052007.545045.433446.894226.0936.52.56.49
VanillaAgent Base Model=GLM-4.6V-Flash2026.0518227.246228.6512425.9148319.5428.67
VanillaAgent Base Model=Qwen3-VL-4B2026.0512420.586025.76027.342427.36.5225.23