Dataset Viewer
Auto-converted to Parquet
image
imagewidth (px)
1.02k
1.02k
elo
int64
957
1.49k
__index_level_0__
int64
30
14.6k
category
stringclasses
12 values
subcategory
stringclasses
121 values
1,489
4,104
Painting
Fresco
1,459
3,124
Photographic
Landscape
1,444
14,433
Illustration
Advertising
1,444
4,104
Painting
Fresco
1,430
2,996
Manga
iconic
1,430
6,826
Fantasy art
painterly
1,428
4,841
3D Model
highly detailed
1,420
13,896
Illustration
Scientific
1,418
6,789
Fantasy art
painterly
1,410
6,399
Manga
vibrant
1,408
10,549
Manga
Japanese comic style
1,408
660
Fantasy art
celestial
1,405
7,448
Fantasy art
majestic
1,402
13,520
Neonpunk
stunningly beautiful
1,402
7,971
Animation
Cutout
1,398
11,855
Illustration
Digital
1,397
4,616
Illustration
Advertising
1,396
3,939
Pixel art
blocky
1,396
8,560
Manga
iconic
1,391
11,507
Digital art
matte painting
1,389
11,221
Fantasy art
fantasy art
1,389
3,655
Painting
Expressionism
1,387
9,345
Manga
vibrant
1,384
13,685
Fantasy art
celestial
1,382
13,613
Manga
vibrant
1,381
822
Cinematic
cinemascope
1,377
3,238
Neonpunk
professional
1,375
12,911
Digital art
matte painting
1,375
12,974
Anime
vibrant
1,375
13,073
Photographic
Fomapan
1,375
2,882
Animation
Animation
1,375
7,448
Fantasy art
majestic
1,373
11,525
Illustration
Manga
1,372
11,787
Pixel art
8-bit graphics
1,372
5,813
Animation
Stop motion
1,372
9,841
Animation
Animation
1,371
5,532
3D Model
highly detailed
1,367
8,490
3D Model
highly detailed
1,366
11,336
Cinematic
moody
1,366
8,039
Digital art
matte painting
1,365
12,320
Illustration
Fashion
1,363
1,252
Digital art
highly detailed
1,363
175
Digital art
illustrative
1,363
4,738
Cinematic
vignette
1,363
6,625
Anime
anime style
1,363
6,812
Fantasy art
majestic
1,363
9,747
Digital art
matte painting
1,362
6,135
3D Model
octane render
1,362
8,583
Photographic
Gold
1,361
10,453
Digital art
matte painting
1,360
11,019
Neonpunk
ultra detailed
1,360
1,605
Illustration
Scientific
1,359
2,172
Animation
Stop motion
1,359
4,443
Anime
studio anime
1,359
6,327
Pixel art
8-bit graphics
1,356
3,275
Painting
Impressionism
1,355
3,656
Pixel art
8-bit graphics
1,354
12,333
Digital art
matte painting
1,354
8,555
Illustration
Manga
1,353
5,475
Animation
Pixel Art
1,352
12,002
Neonpunk
vaporwave
1,352
6,870
Anime
studio anime
1,351
9,854
Fantasy art
ethereal
1,351
5,150
Manga
vibrant
1,349
1,671
Illustration
Fantasy
1,348
10,878
Fantasy art
ethereal
1,348
2,889
Photographic
StreetPan
1,348
9,624
3D Model
volumetric
1,347
13,134
Illustration
Fantasy
1,347
11,272
Painting
Baroque
1,347
6,898
Digital art
painterly
1,346
8,071
Digital art
highly detailed
1,345
10,888
Painting
Digital
1,343
12,412
Painting
Fresco
1,341
14,208
Photographic
Ektachrome
1,341
6,551
Cinematic
bokeh
1,341
1,127
Manga
iconic
1,340
2,296
Painting
Romanticism
1,340
8,890
Anime
vibrant
1,339
4,489
Neonpunk
ultra detailed
1,338
11,725
Photographic
T-Max
1,338
4,833
Anime
studio anime
1,338
761
Animation
Claymation
1,338
8,284
Digital art
painterly
1,337
5,467
Illustration
Fantasy
1,337
7,735
Photographic
CineStill 800T
1,337
9,810
Illustration
Digital
1,336
12,720
Fantasy art
painterly
1,336
3,134
Manga
vibrant
1,336
3,548
Painting
Minimalism
1,335
13,349
Illustration
Digital
1,334
5,037
3D Model
dramatic lighting
1,334
7,106
Illustration
Digital
1,334
13,200
Pixel art
blocky
1,334
7,858
Cinematic
gorgeous
1,333
14,329
Anime
anime style
1,333
9,989
Photographic
Fomapan
1,333
5,046
3D Model
highly detailed
1,333
6,861
Fantasy art
magnificent
1,332
126
Painting
Mural
End of preview. Expand in Data Studio

2k Ranked Images

This dataset contains roughly two thousand images ranked from most preferred to least preferred based on human feedback on pairwise comparisons (>25k responses). The generated images, which are a sample from the open-image-preferences-v1 dataset from the team @data-is-better-together, are rated purely based on aesthetic preference, disregarding the prompt used for generation. We provide the categories of the original dataset for easy filtering.

This is a new modality of preference data for us, different from our typical pairwise preferences, that we wanted to try out. If this type of data is useful of interesting to you, please ❤️ the dataset. If there is community interest and this dataset gets 30 ❤️s we will rank the full set of 17k images!

Examples

Some of the most preferred images

A lot of the most preferred images represent beautiful and serene landscapes and nature scenes.

Some of the least preferred images

The least preferres images are more diverse, but often contain scary, malformed, or just plain boring subjects.

Dataset visualization

Rank your own images through our API!

Downloads last month
50