Compl-AI Board: leaderboard
EU AI Act compliance evaluation: measures LLMs across 27 dimensions including bias (BBQ), toxicity, robustness, calibration, and safety using the Compl-AI framework.
Metric: Average Compliance Score (%). Source: huggingface.co. Saturation forecast: Estimated already saturated. 15 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-4 Preview (1106) | 86.44 | #418 |
| 2 | Claude 3 Opus | 84.9 | #571 |
| 3 | Gemini 1.5 Flash (001) | 80.44 | #577 |
| 4 | GPT-3.5 Turbo (0125) | 76.99 | #984 |
| 5 | Yi Chat 34B | 71.94 | #976 |
| 6 | Qwen 1.5 72B Chat | 71.52 | #776 |
| 7 | Bielik-11B-v2.3-Instruct | 71.12 | #693 |
| 8 | Llama 2 70B Chat (HF) | 70.1 | #1264 |
| 9 | Mixtral 8x7B Instruct (v0.1) | 69.91 | #1073 |
| 10 | Mistral 7B Instruct (v0.3) | 67.96 | #1274 |
| 11 | Mistral 7B Instruct (v0.2) | 67.1 | #1333 |
| 12 | Llama 2 13B Chat Base | 66.13 | #1422 |
| 13 | Mistral-7B-v0.3 | 65.67 | #1355 |
| 14 | Llama 2 7B Chat | 63.09 | #1499 |
| 15 | Gemma 2 9B | 57.99 | #958 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=compl-ai-board · How It Works · Data refreshed daily, snapshot 2026-10-20.