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

OmniBench dataset and benchmarks · SOTA2 Research

Research Datasets

OmniBench

Benchmarks

Task NameDataset NameSOTA MethodSort

Task NameDataset NameSOTA ResultTrendResults

Vision-Audio-Text

Omnibench

47.6Accuracy

34

Audio-Visual Reasoning

OmniBench

63.6Accuracy

16

Humanoid Motion Tracking

OmniBench Walk (Slow)

100Success Rate (SR)

14

OmniBench Low 1.0 (test)

100Success Rate (SR)

14

OmniBench Medium 1.0 (test)

100Success Rate (SR)

14

OmniBench High 1.0 (test)

100Success Rate (SR)

14

Omni-modal reasoning

OmniBench (test)

43.6Score

12

Safety Evaluation

OmniBench

42.47Accuracy

12

Integrated Multimodal Reasoning

OmniBench (held-out)

55.2Causal Score

11

Audio-video understanding

OmniBench

49.1Score

10

Multimodal Reasoning

OmniBench

54.9Causal Score

9

Omni-modal Understanding

OmniBench

58.41Overall Score

8

Humanoid Motion Tracking

OmniBench Jump (Low)

100Success Rate

7

Humanoid Motion Tracking

OmniBench Jump Medium

100Success Rate

7

Humanoid Motion Tracking

OmniBench Jump (High)

90Success Rate (SR)

7

Humanoid Motion Tracking

OmniBench Run Medium

100Success Rate (SR)

7

Humanoid Motion Tracking

OmniBench Run (Fast)

100Success Rate (SR)

7

Humanoid Motion Tracking

OmniBench Walk (Fast)

100Success Rate

7

OmniBench Low 1.0 (test)

100Success Rate (SR)

7

OmniBench Medium 1.0 (test)

100SR

7

OmniBench High 1.0 (test)

100Success Rate (SR)

7

Audio-Visual QA

OmniBench

48.25Accuracy

6

Audio-visual Understanding

OmniBench

56.7Accuracy

5

Image Question Answering

OmniBench

—Primary metric

0

Popular tasks

Humanoid motion tracking Loco-Manip Squat Multimodal Reasoning Safety Evaluation

Follow for updates

Join SOTA2 Contact