Skip to content

Benchmark

Vals Code Migration

Coding benchmarks compare models on programming tasks. Scores are percentages.

Models measured
76
Scores recorded
76
Sources
1
IQ weight
2.05

Score distribution

32%
Mean
32.5%
Median
18.8%
Std dev
6.8%
10th percentile
57.5%
90th percentile
0.314%
Min
0.314%69.8%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 2.05 for chat models, measured over 76 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

Showing 51–76 of 76 results.