Categories Companies Research

Loading the SOTA2 catalog…

Explore

Categories Companies Research

Contribute

Join SOTA2 Contact

SOTA2

© 2026 SOTA2

Build in Bellevue, USA

WizWand is now SOTA2 Research

Papers Benchmarks Datasets Tasks

Large Vision-Language Model Evaluation research benchmarks · SOTA2 Research

Large Vision-Language Model Evaluation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

Unweighted Average

9.8ECE

29

May 12, 2026

LLaVA Evaluation Suite (MMBench, MME, MM-Vet, ScienceQA) 1.5 (test val)

68.5MMBench

16

Feb 26, 2026

29.68MCQ Score

14

Mar 10, 2026

DRBench S Subset

47.22MCQ Accuracy

14

Mar 10, 2026

27.04MCQ Score

14

Mar 10, 2026

72.5Overall Score

13

Feb 26, 2026

1,711.44Perception Score

12

Apr 15, 2026

MMBench v1.0 (dev)

77.6Score

11

Feb 26, 2026

77.8Score

10

Feb 26, 2026

Popular datasets

DRBench LLaVA MMBench MME SEED

Follow for updates

Join SOTA2 Contact