Rank | Creator | model | Bradley-Terry | Elo | Wins | Matches |
|---|---|---|---|---|---|---|
1 |  OpenAI | 4o | 1235.14 | 1075.94 | 117980 | 211946 |
2 |  Black Forest Labs | Flux 1 Pro | 1123.30 | 1041.58 | 915513 | 1704648 |
3 | Black Forest Labs | Flux 1.1 Pro | 1072.92 | 1024.82 | 570595 | 1101778 |
4 |  Ideogram | Ideogram | 1030.56 | 1010.21 | 108790 | 214931 |
5 |  Recraft | Recraft V2 | 1019.76 | 1006.38 | 523519 | 1031617 |
6 |  OpenGVLab | Lumina | 1019.64 | 1006.42 | 157506 | 309612 |
7 |  Google | Imagen 3 | 1005.10 | 1001.05 | 538096 | 1088742 |
8 |  xAI | Aurora | 997.67 | 998.36 | 348588 | 700203 |
9 |  Runway | Frames | 967.23 | 987.08 | 296249 | 604440 |
10 | OpenAI | DALL-E 3 | 953.34 | 981.78 | 764397 | 1563381 |
11 |  Stability.ai | Stable Diffusion 3 | 933.58 | 974.16 | 613529 | 1266304 |
12 |  Midjourney | Midjourney 5.2 | 892.96 | 957.96 | 678941 | 1446462 |
13 |  DeepSeek | Janus 7B | 835.89 | 934.26 | 96128 | 215598 |
What is "Bradley-Terry"?
The Bradley-Terry ranking model is a probabilistic model used to predict outcomes in pairwise comparisons. It assigns a strength parameter (reported score) to each item, indicating its likelihood of winning against another. See the wikipedia article for mathematical details.
What do we consider as "Overall preference"?
Here we evaluate the model across all criteria and determine which model has the best overall performance.
All results are directly based on feedback from real human raters. The process of how we came out with results is best described in our blog post.
