EuroLLM-22B-Preview: benchmark results
Provider: EuroLLM. Released 2025-09-01. Access: Open.
Aggregate Rating 1483 ± 31, rank #2627 of 10916 rated models, from 239 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Polish NLU - PoQuAD | 63.69 | Character F1 (%) | 97.4 |
| EuroEval Latvian NLU - MultiWikiQA LV | 70.34 | Character F1 (%) | 95.7 |
| EuroEval Czech NLU - CSFD Sentiment | 66.13 | Matthews correlation coefficient (%) | 95.6 |
| EuroEval Finnish NLU - Tydiqa FI | 75.96 | Character F1 (%) | 95.4 |
| EuroEval German NLU - Germanquad | 68.1 | Character F1 (%) | 94.7 |
| EuroEval Lithuanian NLU - MultiWikiQA LT | 71.4 | Character F1 (%) | 94.3 |
| EuroEval Dutch NLU - SQuAD NL | 80.69 | Character F1 (%) | 93.8 |
| EuroEval Serbian NLU - MultiWikiQA SR | 65.28 | Character F1 (%) | 93.6 |
| EuroEval Spanish NLU - MLQA ES | 68.53 | Character F1 (%) | 93.6 |
| EuroEval French NLU - FQuAD | 74.47 | Character F1 (%) | 92.9 |
| EuroEval Portuguese NLU - MultiWikiQA PT | 77 | Character F1 (%) | 92.4 |
| EuroEval Greek NLU - MultiWikiQA EL | 71.02 | Character F1 (%) | 92.1 |
Interactive version: theaggregate.ai/model?slug=eurollm-22b-preview · How It Works · Data refreshed daily, snapshot 2026-11-08.