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

Medical Visual Question Answering research benchmarks – Page 4 · SOTA2 Research

Medical Visual Question Answering

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

68.6Average Score

13

Mar 27, 2026

Medical VQA SLAKE, VQA-RAD, VQA-Med 2019/2020/2021

43.42Accuracy

13

Feb 26, 2026

78.8Closed Accuracy

12

Jun 19, 2026

55.11BertScore

12

Apr 28, 2026

Path-VQA (test)

63.61BLEU-1

12

Mar 4, 2026

MedXpert (test)

22.3Accuracy

12

Mar 4, 2026

92Closed Accuracy

11

Apr 21, 2026

66.32Accuracy

11

Apr 3, 2026

73.74Accuracy

11

Apr 3, 2026

Harvard-FairVLMed

88.43Accuracy

11

Apr 3, 2026

Aggregate (SLAKE, VQA-RAD, PathVQA, PMC-VQA) Average (test)

81.2Accuracy

11

Mar 25, 2026

PMC-VQA OOD (test)

0.058ECE

10

Jun 26, 2026

Average (OmniMedVQA, PMC-VQA, MedXpertQA) cross-dataset (test)

0.155ECE

10

Jun 26, 2026

56.47MMMU

10

Apr 28, 2026

VQA-RAD cross-domain

0.789Accuracy

10

Feb 26, 2026

VQA-RAD (in-domain)

83.3Accuracy

10

Feb 26, 2026

89.9Closed-Ended Accuracy

9

Apr 8, 2026

76.6AUROC

9

Apr 6, 2026

VinDr Visual Perturbation (test)

70.1VPT Score

9

Mar 17, 2026

MIMIC Visual Perturbation (test)

73.4VPT Score

9

Mar 17, 2026

VinDr Textual Perturbation (test)

77Anatomy Accuracy

9

Mar 17, 2026

MIMIC Textual Perturbation (test)

79.2Anatomy Accuracy

9

Mar 17, 2026

OmniMed OOD (test)

82.8Accuracy

9

Mar 11, 2026

Rform + Rexact + RDTW

73.3Accuracy

9

May 7, 2026

DeepTumorVQA refined 2025b

3DMedAgent (GPT-5)

66Total Average

9

Feb 26, 2026

Page 4 of 8

Popular datasets

SLAKE VQA-RAD PMC-VQA VQA-Med PathVQA

Follow for updates

Join SOTA2 Contact