CyberGym — leaderboard

Cybersecurity agent benchmark for discovering, exploiting, and reasoning about vulnerabilities in controlled challenge environments.

Metric: Score (self-reported). Source: benchmarklist.com. Status: saturation imminent. 8 models tracked.

Top models

#ModelScore
1Claude Mythos 583.8
2Claude Mythos Preview83.1
3GPT-5.581.8
4GPT-5.479
5Claude Opus 4.878.1
6Claude Opus 4.673.8
7Claude Opus 4.773.1
8Claude Sonnet 4.665.2

Interactive version: theaggregate.ai/benchmark?slug=cybergym · How the rankings work · Data refreshed daily, snapshot 2026-07-22.