Benchmark
Design Arena Elo (models/imagetoimage)
People compare model answers and choose which they prefer. Higher scores are better.
- Models measured
- 25
- Scores recorded
- 28
- Sources
- 1
- IQ weight
- 1.62
Score distribution
- 1,190
- Mean
- 1,215
- Median
- 91.2
- Std dev
- 1,089
- 10th percentile
- 1,277
- 90th percentile
- 995
- Min
This benchmark has a weight of 1.62 for image-generation models, measured over 22 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.