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

MMT-Bench

Benchmarks

Task NameDataset NameSOTA ResultTrend
Multimodal UnderstandingMMT-Bench
Accuracy59.2
25
Multimodal ReasoningMMT-Bench
Accuracy57.88
23
Multi-image UnderstandingMMT-Bench (val)
Score71.8
23
Multimodal EvaluationMMT-Bench
Accuracy62.65
13
Multimodal tasksMMT-Bench 1.0 (test)
Overall63.4
13
General VQAMMT-Bench (val)
Score67
7
Showing 6 of 6 rows