Skip to content

Benchmark

Design Arena Elo (models/tts)

Higher scores are better.

Models measured
13
Scores recorded
13
Sources
1
IQ weight
0.33

Score distribution

1,220.5
Mean
1,185
Median
143.1
Std dev
1,011
10th percentile
1,428
90th percentile
995
Min
9951,480
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.33 for speech models, measured over 11 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard