HELM AIR-Bench 2024 - #39-40.55: Classification of individuals - Health conditions: leaderboard
HELM AIR-Bench 2024 (v1.19.0) refusal rate for level-4 risk category "#39-40.55: Classification of individuals - Health conditions" (level-3 "#39-40: Discrimination/Protected Characteristics Combinations", level-2 "Discrimination/Bias"). Only 30 prompts, so scores move in coarse steps. A GPT-4o judge scores each response from compliance (0) to refusal (1); shown as 0-100, higher means more refusals (safer).
Metric: Refusal Rate (%). Source: crfm.stanford.edu. Saturation forecast: Estimated already saturated. 87 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5 | 100 | #105 |
| 2 | Claude Sonnet 4.5 | 100 | #148 |
| 3 | GPT-4.1 Mini | 100 | #363 |
| 4 | Llama 3.1 70B Instruct | 100 | #567 |
| 5 | GPT-OSS-120B | 100 | #346 |
| 6 | Claude 3.5 Sonnet (20241022) | 100 | #302 |
| 7 | DeepSeek R1 | 100 | #262 |
| 8 | Claude 3.5 Sonnet (20240620) | 100 | #308 |
| 9 | GPT-OSS-20B | 100 | #523 |
| 10 | Gemini 2.5 Pro (Preview 03-25) | 100 | #126 |
| 11 | Mistral 7B Instruct (v0.3) | 100 | #1255 |
| 12 | GPT-4o (2024-08-06) | 100 | #343 |
| 13 | GPT-4 Turbo | 100 | #417 |
| 14 | Claude Haiku 4.5 (20251001) | 100 | #264 |
| 15 | Qwen 2 72B Instruct | 100 | #579 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=helm-air-bench-2024-39-40-55-classification-of-individuals-health-conditions · How It Works · Data refreshed daily, snapshot 2026-11-05.