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

Text-to-Image Generation research benchmarks – Page 36 · SOTA2 Research

Text-to-Image Generation

Benchmarks

Dataset NameSOTA MethodSort

Dataset NameSOTA methodMetricTrendResultsLast Updated

DPG 1024x1024 resolution

78.8DPG Score

3

Mar 31, 2026

SD large 3.5 8B (test)

1.254Image Reward

3

Mar 26, 2026

FLUX.1 dev 12B (test)

1.258Image Reward

3

Mar 26, 2026

ShareGPT-4o-Image SD3.5-Large

35.976CLIP Score

3

Mar 26, 2026

User Study SD 3

54.17Preference Rate

3

Mar 24, 2026

User Study SD 2.1

76.67Preference Rate

3

Mar 24, 2026

T2I-CompBench Color ++

0.621M-LPIPS

3

Mar 24, 2026

76.7Property Score

3

Mar 20, 2026

RareBench 1.0 (test)

81.9Property Score

3

Mar 20, 2026

T2ICompBench Complex category

1.38ImageReward (A-A)

3

Mar 17, 2026

TDM-R1-Zimage (Ours)

0.77GenEval Score

3

Mar 10, 2026

Stable Diffusion 2.1

20.77FID

3

Mar 10, 2026

25.88FID

3

Mar 6, 2026

4.25FID

3

Mar 4, 2026

31.8CLIP-T

3

Mar 4, 2026

Internal Dataset

Unified Latents

4.1gFID@30K

3

Feb 26, 2026

0.307CLIP_L

3

Feb 26, 2026

Amazon Mechanical Turk User Study (test)

43.45Image Fidelity

3

Feb 26, 2026

126.3FID

3

Feb 26, 2026

0.873ImageReward

3

Feb 26, 2026

4.51FID

3

Feb 26, 2026

13.04FID

3

Feb 26, 2026

pokemon-blip-captions v1 (test)

8.32Quality

3

Feb 26, 2026

31.439CS

3

Feb 26, 2026

83TIFA Score

3

Feb 26, 2026

Page 36 of 39

Popular datasets

GenEval MS-COCO DrawBench COCO T2I-CompBench

Follow for updates

Join SOTA2 Contact