Grip on LLMs - Dutch Factuality Average: leaderboard
Metric: Accuracy (%; mean of three GP-IRT estimates, Dutch machine translation). Source: arxiv.org. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | GPT-5 | 76 |
| 2 | Mistral Medium 3 | 75 |
| 3 | GPT-4o | 73 |
| 4 | Qwen 3 32B | 72 |
| 5 | Mistral Large 3 | 71 |
| 6 | Mistral Small 3 | 70 |
| 7 | GPT-5 Mini | 69 |
| 8 | Gemma 3 27B (IT) | 68 |
| 9 | DeepSeek R1 Distill Qwen 32B | 67 |
| 10 | GPT-4o Mini | 65 |
| 11 | aya-expanse-32B | 62 |
| 12 | Gemma 3 12B (IT) | 61 |
| 13 | GPT-5 Nano | 61 |
| 14 | Qwen 3 8B | 59 |
| 15 | GPT-OSS-20B | 57 |
Interactive version: theaggregate.ai/benchmark?slug=grip-on-llms-dutch-factuality-average · How It Works · Data refreshed daily, snapshot 2026-09-19.