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

WebVoyager

Benchmarks

Task NameDataset NameSOTA ResultTrend
Browser-useWebVoyager
Success Rate90.6
14
Web NavigationWebVoyager (test)
Success Rate87
14
GUI NavigationWebVoyager
Success Rate (Allrecipes)88.89
12
Web NavigationWebVoyager 1.0 (test)
Allrecipes62.5
12
Web NavigationWebVoyager
Success Rate88.8
9
Web NavigationWebVoyager
Spearman Correlation0.15
8
Web Navigation and Task CompletionWebVoyager Live Websites
Success Rate (All Rec)38.7
7
Web Browsing TaskWebVoyager Easy
Kimi K2 Success Rate84.3
5
Browser UseWebVoyager (test)
Success Rate87.2
4
Autonomous Web NavigationWebVoyager
Accuracy87
3
Agent Trajectory Verification AgreementWebVoyager
Unterminated Rate (%)3.4
2
Showing 11 of 11 rows