Skip to content

Benchmark

Arena ELO Vision - Humor

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

Models measured
90
Scores recorded
102
Sources
1
IQ weight
1.76

Score distribution

1,244.3
Mean
1,252.6
Median
54.3
Std dev
1,167.5
10th percentile
1,313
90th percentile
1,100.3
Min
1,100.31,347.9
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.76 for chat models, measured over 90 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
101Llama 4 Scout· LMArena data1,102.2
102Llama 4 Maverick· LMArena data1,100.3

Showing 101–102 of 102 results.

Data credits: LMArena, CC BY 4.0 (licence)