Skip to content

Benchmark

Design Arena Elo (models/gamedev)

People compare model answers and choose which they prefer. Higher scores are better.

Models measured
148
Scores recorded
160
Sources
1
IQ weight
1.91

Score distribution

1,136.4
Mean
1,134
Median
150.8
Std dev
942
10th percentile
1,317
90th percentile
745
Min
7451,540
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.91 for chat models, measured over 148 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Showing 151–160 of 160 results.