EuroEval Faroese NLU - FoQA: leaderboard
EuroEval's Faroese natural-language-understanding score — aggregate over the EuroEval Faroese understanding tasks (reading comprehension, linguistic acceptability, NER), from the EuroEval European language model benchmark suite.
Metric: Character F1 (%). Source: euroeval.com. Saturation forecast: Around January 2027. 393 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Qwen3.5-4B-Base | 82.59 | #798 |
| 2 | Mistral Small 3.2 | 79.92 | #605 |
| 3 | Llama 3.1 405B Instruct FP8 | 79.74 | #389 |
| 4 | Magistral Small 1.2 | 79.56 | #570 |
| 5 | Mistral Small 3.1 | 79.48 | #610 |
| 6 | Llama 3.1 70B | 79.34 | #592 |
| 7 | Ministral-3-14B-Reasoning-2512 | 79.31 | #540 |
| 8 | Ministral-3-14B-Instruct-2512 | 78.86 | #615 |
| 9 | Gemma 2 9B (IT) | 78.56 | #788 |
| 10 | gemma-3-12B-pt | 77.27 | #686 |
| 11 | gemma-3-27B-pt | 77.23 | #502 |
| 12 | Gemini 2.5 Flash Lite (Non-reasoning) | 76.6 | #422 (Gemini 2.5 Flash Lite) |
| 13 | GPT-4.1 | 76.47 | #248 |
| 14 | Grok 4.1 Fast (Reasoning) | 76.19 | #217 (Grok 4.1 Fast) |
| 15 | cogito-v1-preview-qwen-14B | 76.19 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=euroeval-faroese-nlu-foqa · How It Works · Data refreshed daily, snapshot 2026-10-20.