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

MMMU

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal UnderstandingMMMU
Accuracy81.8
437
Multi-discipline Multimodal UnderstandingMMMU
Accuracy84.2
363
Multimodal UnderstandingMMMU
MMMU Score67.8
232
Massive Multi-discipline Multimodal UnderstandingMMMU
Accuracy65.5
216
Multi-discipline Multimodal UnderstandingMMMU (val)
Accuracy81.7
212
Multimodal ReasoningMMMU
Accuracy83.89
208
Multimodal UnderstandingMMMU (val)
MMMU Score85.2
199
Multimodal ReasoningMMMU (val)
Accuracy78.2
168
Multimodal ReasoningMMMU Pro
Accuracy85.6
146
Multimodal UnderstandingMMMU (test)
MMMU Score69.6
112
Multimodal UnderstandingMMMU
MMMU Score81.8
102
Multi-modal Question AnsweringMMMU
Accuracy82.3
83
Multimodal UnderstandingMMMU
Accuracy59.63
76
Multimodal UnderstandingMMMU
MMMU Score60.74
69
Video reasoningVideo-MMMU
Accuracy84.6
68
Multi-discipline Multimodal UnderstandingMMMU Pro
Accuracy67.3
66
Vision UnderstandingMMMU
Accuracy72.9
65
Visual Question AnsweringMMMU
Accuracy81.7
54
Multimodal UnderstandingMMMU
Accuracy (MMMU)58
52
Multi-agent discussion attackMMMU
Delta Accuracy2.3
48
General ReasoningMMMU
Overall Score75.4
48
Multimodal ReasoningMMMU
Accuracy85.79
40
Medical Visual Question AnsweringMMMU Health & Medicine (test)
Accuracy74.5
39
Multimodal UnderstandingMMMU
Accuracy56.8
38
Multi-discipline reasoningMMMU (val)
Accuracy81.8
38
Showing 25 of 179 rows
...