external benchmark
SEC-Bench Pro leaderboard
Every model the catalog carries a published SEC-Bench Pro value for, ranked by that value.
Categoryexternal
MeasureSuccess rate
TasksSecurity engineering tasks
DifficultyAdvanced cybersecurity
Published byIntroducing GPT-5.6
Cybersecurity benchmark for agentic vulnerability analysis and exploit-oriented security tasks.
SEC-Bench Pro ranking
4 models with a published SEC-Bench Pro value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Success rate |
|---|---|---|---|
| 1 |  GPT-6 Astra | OpenAI | 85.4 |
| 2 | GPT-5.6 Sol | OpenAI | 71.2 |
| 3 |  DeepSeek V4.1 Flash | DeepSeek | 62.8 |
| 4 | GPT-5.5 | OpenAI | 45.8 |
Evidence key: Observed
Rows are ordered by the value Introducing GPT-5.6 published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.