LLM Trustworthy - Stereotype: leaderboard
Metric: Trust Score (%). Source: huggingface.co. 26 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | gemma-7B (IT) | 100 |
| 2 | Claude 2 | 100 |
| 3 | GPT-4o (2024-05-13) | 99.67 |
| 4 | Llama 3 8B Instruct | 98.33 |
| 5 | Gemini 1.0 Pro | 98.33 |
| 6 | Llama 2 7B Chat | 97.6 |
| 7 | tulu-2-7B | 96.6 |
| 8 | zephyr-7B-beta | 92.6 |
| 9 | tulu-2-13B | 89.33 |
| 10 | GPT-4o Mini (2024-07-18) | 87.34 |
| 11 | falcon-7B Instruct | 87 |
| 12 | GPT-3.5 Turbo (0301) | 87 |
| 13 | vicuna-7B-v1.3 | 81 |
| 14 | Mistral-7B-OpenOrca | 79.33 |
| 15 | GPT-4 (0314) | 77 |
Interactive version: theaggregate.ai/benchmark?slug=llm-trustworthy-stereotype · How It Works · Data refreshed daily, snapshot 2026-09-05.