engines: V8,SpiderMonkey · release: May 2026 · vulnerabilities: 183
OpenAI ↗Best LLMs on SEC-Bench Pro
SEC-Bench ProFinds historical vulnerabilities in V8 or SpiderMonkey and writes a program that triggers the target flaw.
Updated: Data as of:
| Model / creator | Version | Reported score |
|---|---|---|
| May 2026 / GPT-6.2 system card | 88.34% | |
| May 2026 / GPT-6.2 system card | 86.75% |
View all 2 published reports
GPT-6.2 Sol88.34%
May 2026 / GPT-6.2 system cardGPT-6.2 Astra86.75%
May 2026 / GPT-6.2 system cardengines: V8,SpiderMonkey · release: May 2026 · vulnerabilities: 183
OpenAI ↗