EuroEval Slovene NLU - SSJ500K NER: leaderboard

The "SSJ500K NER" subtask of EuroEval Slovene NLU. Slovene-language EuroEval natural-language-understanding suite, separating NLU task performance from the broader all-task EuroEval aggregate.

Metric: Micro F1 excluding MISC (%). Source: euroeval.com. Saturation forecast: Around January 2027. 204 models tracked.

Top models

#ModelScoreOverall rank
1GPT-5.4 Mini (2026-03-17) (High)80.44#169 (GPT-5.4 Mini (2026-03-17))
2Claude Sonnet 4.5 (Thinking)76.38#142 (Claude Sonnet 4.5)
3NVIDIA-Nemotron-3-Nano-30B-A3B-BF1676.38#679
4GPT-5 Nano75.39#425
5GPT-575.31#98
6Claude Sonnet 4.5 (Non-reasoning)74.26#142 (Claude Sonnet 4.5)
7Qwen 3 4B 2507 (Thinking)74.2#538 (Qwen 3 4B 2507)
8Gemma 4 31B (IT)73.49#176
9Gemma 4 26B A4B (IT)72.7#297
10Gemini 3 Flash (Preview) (Non-reasoning)71.77#82 (Gemini 3 Flash (Preview))
11OLMo 3.1 32B (Thinking)71.59#632 (OLMo 3.1 32B)
12Gemini 2.5 Flash (Thinking)71.5#245 (Gemini 2.5 Flash)
13Gemini 2.5 Flash (Non-reasoning)71.18#245 (Gemini 2.5 Flash)
14Grok 4.1 Fast (Reasoning)71.13#217 (Grok 4.1 Fast)
15Gemini 3 Pro (Preview)70.92#66

No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.

Interactive version: theaggregate.ai/benchmark?slug=euroeval-slovene-nlu-ssj500k-ner · How It Works · Data refreshed daily, snapshot 2026-10-20.