Skip to content

Benchmark

Google Real-world Vulnerability Discovery

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

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

Score distribution

71%
Mean
71%
Median
0%
Std dev
71%
10th percentile
71%
90th percentile
71%
Min

Leaderboard

#ModelScore
1Gemini 3.8 Flash Cyber· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim71%

Data credits: Data by LLM Stats