0

ExploitGym

Offensive-cybersecurity benchmark measuring whether a model can turn known software vulnerabilities into working exploits. Companion to ExploitBench.

Domain
cybersecurity
Published
Sep 2026
Notable for
Cybersecurity exploitation benchmark used at the GPT-6 Astra launch.

Cite

Notes

Only stored in your browser.

Attribution

Leaderboard scores
OpenAI
Attribution policy →

Top score 42.4% by GPT-6 Astra - 1 model reporting (1 frontier)

Top models

1
ExploitGymBar chart with 1 bar. Highest value: GPT-6 Astra at 42.4.
1 model

FAQ

What is ExploitGym?
Offensive-cybersecurity benchmark measuring whether a model can turn known software vulnerabilities into working exploits. Companion to ExploitBench.
What is the current top score on ExploitGym?
The top reported score is 42.4% by GPT-6 Astra, across 1 model reporting (1 from frontier labs).