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

Multimodal Question Answering research benchmarks – Page 3 · SOTA2 Research

Multimodal Question Answering

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

MLLM Evaluation Suite (GQA, MMB, MMB-CN, MME, POPE, SQA, VQAv2, VQA-Text, VizWiz) (test)

64.27GQA Accuracy

11

Mar 4, 2026

ScienceQA (SQA)

3.86Avg Accepted Length

10

Feb 26, 2026

MultiModalQA (val)

65.1EM

10

Feb 26, 2026

MLLM Pre-trained Evaluation Suite VQAv2, GQA, VizWiz, SQA, TextVQA, POPE, MM-Bench, MM-Bench-CN

Pre-trained Model

78.5VQAv2 Accuracy

9

Jun 25, 2026

PIVOT Evaluation Suite

LLaVA-Next-Qwen2

62Overall Accuracy

9

Apr 14, 2026

MMBench CN (test)

88.9Accuracy

9

Feb 26, 2026

41.2Score

9

Feb 26, 2026

64.3Score

9

Feb 26, 2026

OpenCLIP + BLIP-2

64.8ROrig@k

8

Mar 4, 2026

Frontier-comparable pool 20-eval macro mean (test)

10.23AME Change (pp)

7

Jun 25, 2026

17.1EM

7

Jun 11, 2026

MLLM Evaluation Suite (HallBench, MME, TextVQA, ChartQA, AI2D, RealWorldQA, CCBench, OCRVQA, SQA-IMG, POPE) (test)

49.8HallBench

7

Apr 10, 2026

LLaVA-NeXT-7B Evaluation Suite

5.6FLOPs (TFLOPs)

6

Jul 1, 2026

79.5Accuracy (%)

6

May 29, 2026

MultiModalQA 95 (test)

82.7Accuracy

6

May 29, 2026

Encyclopedic-VQA 90 (test)

84.3Accuracy

6

May 29, 2026

6.52Accuracy

6

Apr 10, 2026

34.17ROUGE-1

6

Mar 10, 2026

51.11EM

6

Feb 26, 2026

71.9Answer Accuracy

4

Mar 10, 2026

80.6Answer Accuracy

4

Mar 10, 2026

ScienceQA (SQA) (test)

Vanilla (LLaVA-NeXT-7B)

70.2SQA Accuracy

4

Feb 26, 2026

58.2Accuracy

3

Apr 10, 2026

MMQA SemBench 1.0 (test)

23.49Latency (s)

2

Jul 1, 2026

71.17ROUGE-L

2

May 19, 2026

Page 3 of 3

Popular datasets

MMBench WebQA ScienceQA MMQA MultiModalQA

Follow for updates

Join SOTA2 Contact