EuroEval Faroese NLU - FONE — leaderboard

Metric: Named entity recognition Score (%). Source: euroeval.com. 401 models tracked.

Top models

#ModelScore
1Gemma 4 31B (IT)84.74
2Claude Sonnet 4.5 (Thinking)83.21
3Gemma 4 26B A4B (IT)82.95
4Llama 3.1 405B Instruct FP881.57
5Gemini 3 Flash (Preview) (Non-reasoning)81.04
6GPT-5.4 Mini (High)80.76
7GPT-5.4 Mini (Medium)80.16
8Llama 3.1 70B80.08
9Mistral Small 3.179.69
10Claude Sonnet 4.5 (Non-reasoning)79.34
11Qwen 3 4B 2507 (Thinking)79.02
12Mistral Small 3.278.05
13Llama 3.1 405B78.02
14OLMo 3.1 32B (Thinking)77.58
15Llama 4 Scout Base77.46

Interactive version: theaggregate.ai/benchmark?slug=euroeval-faroese-nlu-fone · How the rankings work · Data refreshed daily, snapshot 2026-07-22.