OpenAI GPT-5.6 System Card - Biology Refusal - Benign: leaderboard
Biology Model Refusal Evaluation table of OpenAI's GPT-5.6 System Card (Deployment Safety Hub, 2026-07-09): share of responses to benign biology prompts graded not over-refused, model responses only without monitors. Percent (the card prints 0-1), higher is better. GPT-5.2, 5.4 and 5.5 Thinking and GPT-5.6 Sol, Terra and Luna.
Metric: Not overrefuse rate (%). Source: deploymentsafety.openai.com. Saturation forecast: Estimated already saturated. 6 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.6 Sol | 98.9 | #22 |
| 2 | GPT-5.6 Luna | 98.9 | #79 |
| 3 | GPT-5.6 Terra | 97.8 | #40 |
| 4 | GPT-5.5 | 91.7 | #33 |
| 5 | GPT-5.2 (Thinking) | 84.8 | #115 (GPT-5.2) |
| 6 | GPT-5.4 (Thinking) | 69.6 | #83 (GPT-5.4) |
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-system-card-biology-refusal-benign · How It Works · Data refreshed daily, snapshot 2026-11-05.