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

MMLongBench dataset and benchmarks – Page 2 · SOTA2 Research

Research Datasets

MMLongBench

Benchmarks

Task NameDataset NameSOTA MethodSort

Task NameDataset NameSOTA ResultTrendResults

Long-document Visual Question Answering

MMLongBench 256K context

31.63MMLB-D Score

4

Visual Needle In A Haystack

MMLongBench MM-NIAH Reason.

51.87Accuracy

2

Visual Needle In A Haystack

MMLongBench MM-NIAH Retriev

25.56Accuracy

2

Visual Needle In A Haystack

MMLongBench MM-NIAH Count

38.95Soft Accuracy

2

Visual Needle In A Haystack

MMLongBench VH-Multi

54.77Accuracy

2

Visual Needle In A Haystack

MMLongBench VH-Single

55.85Accuracy

2

Visual Retrieval Augmented Generation

MMLongBench ViQuAE

55.68SubEM

2

Visual Retrieval Augmented Generation

MMLongBench InfoSeek

60.76SubEM

2

Document Question Answering

MMLongBench SlideVQA

68.68Document QA Accuracy

2

Document Question Answering

MMLongBench MMLongDoc

53.29Document QA Accuracy

2

Document Question Answering

MMLongBench LongDocURL

60.6QA Accuracy

2

In-Context Learning

MMLongBench iNat

22Accuracy

2

In-Context Learning

MMLongBench Food

34Accuracy

2

In-Context Learning

MMLongBench Cars

0.264Accuracy

2

Page 2 of 2

Popular tasks

Document Question Answering Long-document Visual Question Answering Visual Needle In A Haystack Retrieval In-Context Learning

Follow for updates

Join SOTA2 Contact