# EuroEval

> The robust European language model benchmark.

EuroEval evaluates language models (encoders, decoders, encoder-decoders, base and instruction-tuned models) across 30+ European languages with reproducible benchmarks. The Python package is at https://github.com/EuroEval/EuroEval.

## Documentation

### About

- [About overview](https://euroeval.com/)

### Leaderboards

- [Leaderboards overview](https://euroeval.com/leaderboards)
- [European](https://euroeval.com/leaderboards/european)
- [Baltic](https://euroeval.com/leaderboards/baltic)
- [Finnic](https://euroeval.com/leaderboards/finnic)
- [Germanic](https://euroeval.com/leaderboards/germanic)
- [Mainland Scandinavian](https://euroeval.com/leaderboards/mainland-scandinavian)
- [Romance](https://euroeval.com/leaderboards/romance)
- [Slavic](https://euroeval.com/leaderboards/slavic)
- [Albanian](https://euroeval.com/leaderboards/albanian)
- [Bosnian](https://euroeval.com/leaderboards/bosnian)
- [Bulgarian](https://euroeval.com/leaderboards/bulgarian)
- [Catalan](https://euroeval.com/leaderboards/catalan)
- [Croatian](https://euroeval.com/leaderboards/croatian)
- [Czech](https://euroeval.com/leaderboards/czech)
- [Danish](https://euroeval.com/leaderboards/danish)
- [Dutch](https://euroeval.com/leaderboards/dutch)
- [English](https://euroeval.com/leaderboards/english)
- [Estonian](https://euroeval.com/leaderboards/estonian)
- [Faroese](https://euroeval.com/leaderboards/faroese)
- [Finnish](https://euroeval.com/leaderboards/finnish)
- [French](https://euroeval.com/leaderboards/french)
- [German](https://euroeval.com/leaderboards/german)
- [Greek](https://euroeval.com/leaderboards/greek)
- [Hungarian](https://euroeval.com/leaderboards/hungarian)
- [Icelandic](https://euroeval.com/leaderboards/icelandic)
- [Italian](https://euroeval.com/leaderboards/italian)
- [Latvian](https://euroeval.com/leaderboards/latvian)
- [Lithuanian](https://euroeval.com/leaderboards/lithuanian)
- [Norwegian](https://euroeval.com/leaderboards/norwegian)
- [Polish](https://euroeval.com/leaderboards/polish)
- [Portuguese](https://euroeval.com/leaderboards/portuguese)
- [Romanian](https://euroeval.com/leaderboards/romanian)
- [Serbian](https://euroeval.com/leaderboards/serbian)
- [Slovak](https://euroeval.com/leaderboards/slovak)
- [Slovene](https://euroeval.com/leaderboards/slovene)
- [Spanish](https://euroeval.com/leaderboards/spanish)
- [Swedish](https://euroeval.com/leaderboards/swedish)
- [Ukrainian](https://euroeval.com/leaderboards/ukrainian)

### Evaluation Queue


### Tasks

- [Tasks overview](https://euroeval.com/tasks)
- [Bias Detection](https://euroeval.com/tasks/bias-detection)
- [Common-sense Reasoning](https://euroeval.com/tasks/common-sense-reasoning)
- [European Values](https://euroeval.com/tasks/european-values)
- [Grammatical Error Detection](https://euroeval.com/tasks/grammatical-error-detection)
- [Instruction-following](https://euroeval.com/tasks/instruction-following)
- [Knowledge](https://euroeval.com/tasks/knowledge)
- [Linguistic Acceptability](https://euroeval.com/tasks/linguistic-acceptability)
- [Named Entity Recognition](https://euroeval.com/tasks/named-entity-recognition)
- [Natural Language Inference](https://euroeval.com/tasks/natural-language-inference)
- [Reading Comprehension](https://euroeval.com/tasks/reading-comprehension)
- [Sentiment Classification](https://euroeval.com/tasks/sentiment-classification)
- [Simplification](https://euroeval.com/tasks/simplification)
- [Speed](https://euroeval.com/tasks/speed)
- [Summarization](https://euroeval.com/tasks/summarization)
- [Tool calling](https://euroeval.com/tasks/tool-calling)
- [Translation](https://euroeval.com/tasks/translation)

### Datasets

- [Datasets overview](https://euroeval.com/datasets)
- [Albanian](https://euroeval.com/datasets/albanian)
- [Belarusian](https://euroeval.com/datasets/belarusian)
- [Bosnian](https://euroeval.com/datasets/bosnian)
- [Bulgarian](https://euroeval.com/datasets/bulgarian)
- [Catalan](https://euroeval.com/datasets/catalan)
- [Croatian](https://euroeval.com/datasets/croatian)
- [Czech](https://euroeval.com/datasets/czech)
- [Danish](https://euroeval.com/datasets/danish)
- [Dutch](https://euroeval.com/datasets/dutch)
- [English](https://euroeval.com/datasets/english)
- [Estonian](https://euroeval.com/datasets/estonian)
- [Faroese](https://euroeval.com/datasets/faroese)
- [Finnish](https://euroeval.com/datasets/finnish)
- [French](https://euroeval.com/datasets/french)
- [German](https://euroeval.com/datasets/german)
- [Greek](https://euroeval.com/datasets/greek)
- [Hungarian](https://euroeval.com/datasets/hungarian)
- [Icelandic](https://euroeval.com/datasets/icelandic)
- [Italian](https://euroeval.com/datasets/italian)
- [Latvian](https://euroeval.com/datasets/latvian)
- [Lithuanian](https://euroeval.com/datasets/lithuanian)
- [Norwegian](https://euroeval.com/datasets/norwegian)
- [Polish](https://euroeval.com/datasets/polish)
- [Portuguese](https://euroeval.com/datasets/portuguese)
- [Romanian](https://euroeval.com/datasets/romanian)
- [Serbian](https://euroeval.com/datasets/serbian)
- [Slovak](https://euroeval.com/datasets/slovak)
- [Slovene](https://euroeval.com/datasets/slovene)
- [Spanish](https://euroeval.com/datasets/spanish)
- [Swedish](https://euroeval.com/datasets/swedish)
- [Ukrainian](https://euroeval.com/datasets/ukrainian)

### Methodology

- [Methodology overview](https://euroeval.com/methodology)

### FAQ

- [FAQ overview](https://euroeval.com/faq)

### Python Package

- [Python Package overview](https://euroeval.com/python-package)

### API Reference

- [API Reference overview](https://euroeval.com/api)

## Leaderboard CSVs

Raw evaluation results in DataWrapper-formatted CSV. Direct download URLs:

- [albanian_all_models.csv](https://euroeval.com/leaderboards-csv/albanian_all_models.csv)
- [albanian_generative.csv](https://euroeval.com/leaderboards-csv/albanian_generative.csv)
- [baltic_all_models.csv](https://euroeval.com/leaderboards-csv/baltic_all_models.csv)
- [baltic_generative.csv](https://euroeval.com/leaderboards-csv/baltic_generative.csv)
- [bosnian_all_models.csv](https://euroeval.com/leaderboards-csv/bosnian_all_models.csv)
- [bosnian_generative.csv](https://euroeval.com/leaderboards-csv/bosnian_generative.csv)
- [bulgarian_all_models.csv](https://euroeval.com/leaderboards-csv/bulgarian_all_models.csv)
- [bulgarian_generative.csv](https://euroeval.com/leaderboards-csv/bulgarian_generative.csv)
- [catalan_all_models.csv](https://euroeval.com/leaderboards-csv/catalan_all_models.csv)
- [catalan_generative.csv](https://euroeval.com/leaderboards-csv/catalan_generative.csv)
- [croatian_all_models.csv](https://euroeval.com/leaderboards-csv/croatian_all_models.csv)
- [croatian_generative.csv](https://euroeval.com/leaderboards-csv/croatian_generative.csv)
- [czech_all_models.csv](https://euroeval.com/leaderboards-csv/czech_all_models.csv)
- [czech_generative.csv](https://euroeval.com/leaderboards-csv/czech_generative.csv)
- [danish_all_models.csv](https://euroeval.com/leaderboards-csv/danish_all_models.csv)
- [danish_generative.csv](https://euroeval.com/leaderboards-csv/danish_generative.csv)
- [dutch_all_models.csv](https://euroeval.com/leaderboards-csv/dutch_all_models.csv)
- [dutch_generative.csv](https://euroeval.com/leaderboards-csv/dutch_generative.csv)
- [english_all_models.csv](https://euroeval.com/leaderboards-csv/english_all_models.csv)
- [english_generative.csv](https://euroeval.com/leaderboards-csv/english_generative.csv)
- [estonian_all_models.csv](https://euroeval.com/leaderboards-csv/estonian_all_models.csv)
- [estonian_generative.csv](https://euroeval.com/leaderboards-csv/estonian_generative.csv)
- [european_all_models.csv](https://euroeval.com/leaderboards-csv/european_all_models.csv)
- [european_generative.csv](https://euroeval.com/leaderboards-csv/european_generative.csv)
- [faroese_all_models.csv](https://euroeval.com/leaderboards-csv/faroese_all_models.csv)
- [faroese_generative.csv](https://euroeval.com/leaderboards-csv/faroese_generative.csv)
- [finnic_all_models.csv](https://euroeval.com/leaderboards-csv/finnic_all_models.csv)
- [finnic_generative.csv](https://euroeval.com/leaderboards-csv/finnic_generative.csv)
- [finnish_all_models.csv](https://euroeval.com/leaderboards-csv/finnish_all_models.csv)
- [finnish_generative.csv](https://euroeval.com/leaderboards-csv/finnish_generative.csv)
- [french_all_models.csv](https://euroeval.com/leaderboards-csv/french_all_models.csv)
- [french_generative.csv](https://euroeval.com/leaderboards-csv/french_generative.csv)
- [german_all_models.csv](https://euroeval.com/leaderboards-csv/german_all_models.csv)
- [german_generative.csv](https://euroeval.com/leaderboards-csv/german_generative.csv)
- [germanic_all_models.csv](https://euroeval.com/leaderboards-csv/germanic_all_models.csv)
- [germanic_generative.csv](https://euroeval.com/leaderboards-csv/germanic_generative.csv)
- [greek_all_models.csv](https://euroeval.com/leaderboards-csv/greek_all_models.csv)
- [greek_generative.csv](https://euroeval.com/leaderboards-csv/greek_generative.csv)
- [hungarian_all_models.csv](https://euroeval.com/leaderboards-csv/hungarian_all_models.csv)
- [hungarian_generative.csv](https://euroeval.com/leaderboards-csv/hungarian_generative.csv)
- [icelandic_all_models.csv](https://euroeval.com/leaderboards-csv/icelandic_all_models.csv)
- [icelandic_generative.csv](https://euroeval.com/leaderboards-csv/icelandic_generative.csv)
- [italian_all_models.csv](https://euroeval.com/leaderboards-csv/italian_all_models.csv)
- [italian_generative.csv](https://euroeval.com/leaderboards-csv/italian_generative.csv)
- [latvian_all_models.csv](https://euroeval.com/leaderboards-csv/latvian_all_models.csv)
- [latvian_generative.csv](https://euroeval.com/leaderboards-csv/latvian_generative.csv)
- [lithuanian_all_models.csv](https://euroeval.com/leaderboards-csv/lithuanian_all_models.csv)
- [lithuanian_generative.csv](https://euroeval.com/leaderboards-csv/lithuanian_generative.csv)
- [mainland_scandinavian_all_models.csv](https://euroeval.com/leaderboards-csv/mainland_scandinavian_all_models.csv)
- [mainland_scandinavian_generative.csv](https://euroeval.com/leaderboards-csv/mainland_scandinavian_generative.csv)
- [norwegian_all_models.csv](https://euroeval.com/leaderboards-csv/norwegian_all_models.csv)
- [norwegian_generative.csv](https://euroeval.com/leaderboards-csv/norwegian_generative.csv)
- [polish_all_models.csv](https://euroeval.com/leaderboards-csv/polish_all_models.csv)
- [polish_generative.csv](https://euroeval.com/leaderboards-csv/polish_generative.csv)
- [portuguese_all_models.csv](https://euroeval.com/leaderboards-csv/portuguese_all_models.csv)
- [portuguese_generative.csv](https://euroeval.com/leaderboards-csv/portuguese_generative.csv)
- [romance_all_models.csv](https://euroeval.com/leaderboards-csv/romance_all_models.csv)
- [romance_generative.csv](https://euroeval.com/leaderboards-csv/romance_generative.csv)
- [romanian_all_models.csv](https://euroeval.com/leaderboards-csv/romanian_all_models.csv)
- [romanian_generative.csv](https://euroeval.com/leaderboards-csv/romanian_generative.csv)
- [serbian_all_models.csv](https://euroeval.com/leaderboards-csv/serbian_all_models.csv)
- [serbian_generative.csv](https://euroeval.com/leaderboards-csv/serbian_generative.csv)
- [slavic_all_models.csv](https://euroeval.com/leaderboards-csv/slavic_all_models.csv)
- [slavic_generative.csv](https://euroeval.com/leaderboards-csv/slavic_generative.csv)
- [slovak_all_models.csv](https://euroeval.com/leaderboards-csv/slovak_all_models.csv)
- [slovak_generative.csv](https://euroeval.com/leaderboards-csv/slovak_generative.csv)
- [slovene_all_models.csv](https://euroeval.com/leaderboards-csv/slovene_all_models.csv)
- [slovene_generative.csv](https://euroeval.com/leaderboards-csv/slovene_generative.csv)
- [spanish_all_models.csv](https://euroeval.com/leaderboards-csv/spanish_all_models.csv)
- [spanish_generative.csv](https://euroeval.com/leaderboards-csv/spanish_generative.csv)
- [swedish_all_models.csv](https://euroeval.com/leaderboards-csv/swedish_all_models.csv)
- [swedish_generative.csv](https://euroeval.com/leaderboards-csv/swedish_generative.csv)
- [ukrainian_all_models.csv](https://euroeval.com/leaderboards-csv/ukrainian_all_models.csv)
- [ukrainian_generative.csv](https://euroeval.com/leaderboards-csv/ukrainian_generative.csv)
