BrowseComp
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
BrowseComp v1.0 (test)
55Success Rate
23
BrowseComp English
54.9Score
22
BrowseComp V3
68.03Success Rate (SR) - Avg
22
BrowseComp-plus
78.41Accuracy
22
browsecomp
74.5Accuracy
21
BrowseComp-ZH
68.7Score
21
BrowseComp-Plus (BCP) 830-query (evaluation)
23.44Raw KV Reads
20
BrowseComp+ (test)
56.4Accuracy
20
BrowseComp standard (full)
51.5Pass@1
20
BrowseComp
60.6Overall Score
19
BrowseComp
77.3Score
19
BrowseComp (test)
54.9Accuracy
19
BrowseComp-zh (BC-zh)
42.9Pass@1 Accuracy
19
BrowseComp
77.9Score
19
BrowseComp
51.5Success Rate
19
BrowseComp Chinese (full)
58.1Pass@1
19
BrowseComp
85.9Score
18
BrowseComp
84.4Score
18
BrowseComp-ZH
82.4Score
18
BrowseComp-Plus
83.1Accuracy
18
BrowseComp
68.9Avg@3 Score
18
BrowseComp-zh
33.3Pass@1
18
BrowseComp-ZH
58.1Pass@1
18
BrowseComp
50.9Pass@1
18
BrowseComp-zh
66.6Accuracy
18