OpenAI GPT-5.6 August Update System Card - Cyber Ranges: leaderboard
Cyber range table of OpenAI's GPT-5.6 August Update System Card (Deployment Safety Hub, 2026-08-06): combined pass rate over emulated network-attack scenarios, run at maximum reasoning effort. Percent, higher is better. GPT-5.6 Sol (August 2026) has one scenario marked n/a, so its rate covers one scenario fewer than the other models.
Metric: Combined Pass Rate (%). Source: deploymentsafety.openai.com. Saturation forecast: Around February 2027. 5 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 92.3 | #33 |
| 2 | GPT-5.6 Sol (August 2026) | 83.3 | |
| 3 | GPT-5.3 Codex | 76.9 | #72 |
| 4 | GPT-5.4 (Thinking) | 69.2 | #83 (GPT-5.4) |
| 5 | GPT-5.6 Luna (August 2026) | 61.5 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-5-6-august-update-system-card-cyber-ranges · How It Works · Data refreshed daily, snapshot 2026-11-05.