CompliBench - Healthcare - Guideline Accuracy: leaderboard
Metric: Strict guideline accuracy (%): share of compliant assistant turns in the healthcare domain for which the judge predicts the correct governing guideline and no violation, of an LLM judge on CompliBench's synthesized multi-turn dialogues with injected, adversarially optimized guideline violations, mean of four runs at default reasoning effort; higher is better. Source: arxiv.org. Saturation forecast: Estimated already saturated. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 92.86 |
| 2 | Gemini 3 Pro | 92.39 |
| 3 | Qwen 3.5 Plus | 92.39 |
| 4 | GPT-4o | 90.41 |
| 5 | DeepSeek V3.2 (Thinking) | 89.95 |
| 6 | Qwen 3 Max | 88.24 |
| 7 | Claude Sonnet 4.6 | 87.33 |
| 8 | GLM-5 | 86.18 |
| 9 | Qwen 3 30B A3B | 84.55 |
| 10 | Kimi K2.5 | 79.35 |
| 11 | Qwen 3 32B | 62.42 |
| 12 | Qwen 3 4B | 54.43 |
| 13 | GPT-4o Mini | 51.13 |
| 14 | Qwen 3 8B | 49.34 |
| 15 | Qwen 3 14B | 40.49 |
Interactive version: theaggregate.ai/benchmark?slug=complibench-healthcare-guideline-accuracy · How It Works · Data refreshed daily, snapshot 2026-10-07.