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

WebTailBench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Long-horizon web tasksWebTailBench long-horizon tasks
SR@5040.2
34
Browser-useWebTailBench
Success Rate63.5
13
Web NavigationWebTailBench (live)
Accuracy72.2
11
Web Task CompletionWebTailBench v2
Success Rate (Avg)29.5
3
Agent Trajectory Verification AgreementWebTailBench
Unterminated Rate7.7
2
Showing 5 of 5 rows