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

HLE-VL

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal Deep SearchHLE-VL
Accuracy (HLE-VL)19
18
Multimodal long-horizon complex reasoningHLE-VL
Score25.1
13
Multimodal SearchHLE-VL
Score36
8
Vision Agent CapacityHLE-VL
Score39.2
6
Visual SearchHLE-VL
Pass@136
4
Showing 5 of 5 rows