Skip to content

Benchmark

SRE-Bench · 1 Attempt

Scores are percentages. Higher is better.

Models measured
3
Scores recorded
3
Sources
1
IQ weight
No weight published

Score distribution

52.1%
Mean
55.9%
Median
30.9%
Std dev
12.5%
10th percentile
88%
90th percentile
12.5%
Min
12.5%88%
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#ModelScore
1GPT-6 Astra1 attempt· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.88%
2GPT-5.6 Sol1 attempt· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.55.9%
3Claude Opus 51 attempt· Reported by OpenAI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.12.5%

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.