List your company
ResearchFormerly WizWand

Web Task Success on Mind2Web Live

34.7Task Success RatePRO-CUA
16.29221.07125.8530.629May 27, 2026
Updated 5mo ago

Evaluation results

MethodLinks
PRO-CUATraining Paradigm=Self-evolving 4B, Ext. Expert Steps=02026.0534.7
PRO-CUA-8Training Paradigm=Self-evolving 8B, Ext. Expert Steps=02026.0530.6
Rule-based Step-RLTraining Paradigm=Self-evolving 4B, Ext. Expert Steps=02026.0527.8
FBCTraining Paradigm=Self-evolving 4B, Ext. Expert Steps=02026.0526.4
Rule-based Step-RLTraining Paradigm=Self-evolving 8B, Ext. Expert Steps=02026.0525
FBCTraining Paradigm=Self-evolving 8B, Ext. Expert Steps=02026.0523.6
Qwen3-VL-8B-InstructTraining Paradigm=Self-evolving 8B, Ext. Expert Steps=02026.0520.8
WebSTAR-32BTraining Paradigm=External expert / closed data, Ext. Expert Steps=100K2026.0520.4
UI-TARS-1.5-7BTraining Paradigm=External expert / closed data, Ext. Expert Steps=Closed data2026.0518.1
Qwen3-VL-4B-InstructTraining Paradigm=Self-evolving 4B, Ext. Expert Steps=02026.0518.1
WebSTAR-7BTraining Paradigm=External expert / closed data, Ext. Expert Steps=100K2026.0517