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

Visual Question Answering research benchmarks – Page 41 · SOTA2 Research

Visual Question Answering

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

MathVerse vision-only (test)

Proprietary API SOTA (Google, 2024)

54.8Accuracy

4

Feb 26, 2026

MathVista mini (test)

Proprietary API SOTA (SenseTime, 2024)

78.4Accuracy

4

Feb 26, 2026

CLEVR 1.0 (val)

99.7Accuracy

4

Feb 26, 2026

Histopathology Visual Chat (Evaluation Set)

78.4Conv Score

4

Feb 26, 2026

39.2Accuracy

4

Feb 26, 2026

HallusionBench HBI (all)

45.21Score

4

Feb 26, 2026

70.2Mean Accuracy

4

Feb 26, 2026

0.872Accuracy

4

Feb 26, 2026

xGQA yes/no (test)

53.22Accuracy (en)

4

Feb 26, 2026

0.661Inference Time

4

Feb 26, 2026

40.81Accuracy

4

Feb 26, 2026

51.68Accuracy

4

Feb 26, 2026

BizDocs (held-out)

76.4VQA Score

4

Feb 26, 2026

DocVQA (held-out)

82.8ANLS

4

Feb 26, 2026

86.7ANLS

4

Feb 26, 2026

0.546ANLS

4

Feb 26, 2026

PROMPTCAP + GPT-3

53FL

4

Feb 26, 2026

InfographicVQA (val)

54.8ANLS

4

Feb 26, 2026

8.7GFLOPs

4

Feb 26, 2026

FigureQA Original color scheme (test)

94.88Accuracy

4

Feb 26, 2026

Medical VQA DME dataset (test)

Consistency-preserving VQA

0.835Overall Accuracy

4

Feb 26, 2026

69.8Acc-MC

4

Feb 26, 2026

SQOOP Two objects per image (test)

all - sub-task - all

88.5Accuracy

4

Feb 27, 2026

SQOOP Five objects per image (test)

all - sub-task - all

99.96Accuracy

4

Feb 27, 2026

Human annotations

97.5Turing Test Passing Rate

4

Feb 26, 2026

Page 41 of 46

Popular datasets

VQA GQA TextVQA Clevr Vizwiz

Follow for updates

Join SOTA2 Contact