Skip to content

Benchmark

Code Arena

Coding benchmarks compare models on programming tasks. Scores are Elo ratings. Higher ratings are better.

Models measured
1
Scores recorded
1
Sources
1
IQ weight
No weight published

Score distribution

1,588
Mean
1,588
Median
0
Std dev
1,588
10th percentile
1,588
90th percentile
1,588
Min

Leaderboard

#ModelScore
1Gemini 3.7 FlashWeb development· Reported by Google DeepMind· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1,588

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.