Skip to content

Benchmark

Design Arena Elo (agents/agon_godot)

Agentic benchmarks measure how well models carry out tasks that take several steps. Higher scores are better.

Models measured
21
Scores recorded
22
Sources
1
IQ weight
2.37

Score distribution

1,166.3
Mean
1,202
Median
102.7
Std dev
1,026
10th percentile
1,263
90th percentile
927
Min
9271,309
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.37 for chat models, measured over 21 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard