UD Ancient Greek PROIEL
Models analyze Ancient Greek passages from Herodotus and the New Testament. They identify dictionary forms and grammatical relationships between words.
Riemenschneider and Frank report 97.48% lemma accuracy for GreTa with character embeddings on the 1,047 test sentences. The OdyCy authors also evaluated this treebank.
- Made by
- University of Oslo
- Released
- November 2015
- Kind
- Reading in many languages
- Level
- University
- Scored
- Checked automatically
- Answers
- Short answer
- Tasks
- All public
- Measures
- Capability
- Subject
- Languages and literature
- Language
- Ancient Greek
- Interaction
- Single turn
- Items from
- Real work and use, Another benchmark
- Refresh
- New versions replace it
- Run by
- Anyone
- Based on
- A version of another benchmark
HIT-SCIR (Harbin) tops the leaderboard with 79.3%
Best score per entry from the benchmark's own leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 79.3 HIT-SCIR (Harbin)
- 76.8 Stanford (Stanford)
- 76.5 ICS PAS (Warszawa)
- 75.9 TurkuNLP (Turku)
- 75.8 UDPipe Future (Praha)
- 75.0 CEA LIST (Paris)
- 74.5 ParisNLP (Paris)
- 74.5 LATTICE (Paris)
- 73.8 AntNLP (Shanghai)
- 73.0 NLP-Cube (București)
- 72.5 iParse (Pittsburgh)
- 72.2 Uppsala (Uppsala)
- 71.8 IBM NY (Yorktown Heights)
- 71.6 SLT-Interactions (Bengaluru)
- 71.0 LeisureX (Shanghai)
- 69.5 Fudan (Shanghai)
- 67.6 BASELINE UDPipe 1.2 (Praha)
- 67.3 UniMelb (Melbourne)
- 66.1 Phoenix (Shanghai)
- 64.8 BOUN (İstanbul)
- 63.3 CUNI x-ling (Praha)
- 62.1 KParse (İstanbul)
- 58.4 ArmParser (Yerevan)
- 55.5 ONLP lab (Ra'anana)
- 1.89 HUJI (Yerushalayim)
- 0.0 SParse (İstanbul)
- 1 HIT-SCIR (Harbin) software1-P 79.3%
- 2 Stanford (Stanford) software2 76.8%
- 3 ICS PAS (Warszawa) software1-P 76.5%
- 4 TurkuNLP (Turku) software1-P 75.9%
- 5 UDPipe Future (Praha) software1-P 75.8%
- 6 CEA LIST (Paris) software1-P 75%
- 7 ParisNLP (Paris) software1-P 74.5%
- 8 LATTICE (Paris) software1-P 74.5%
- 9 AntNLP (Shanghai) software1-P 73.8%
- 10 NLP-Cube (București) software1-P 73%
- 11 iParse (Pittsburgh) software1-P 72.5%
- 12 Uppsala (Uppsala) software1-P 72.2%
- 13 IBM NY (Yorktown Heights) software1-P 71.8%
- 14 SLT-Interactions (Bengaluru) software2-P 71.6%
- 15 LeisureX (Shanghai) software1-P 71%
- 16 Fudan (Shanghai) software5-P 69.5%
- 17 BASELINE UDPipe 1.2 (Praha) software1-P 67.6%
- 18 UniMelb (Melbourne) software1-P 67.3%
- 19 Phoenix (Shanghai) software1-P 66.1%
- 20 BOUN (İstanbul) software2-P 64.8%
- 21 CUNI x-ling (Praha) software1-P 63.3%
- 22 KParse (İstanbul) software1-P 62.1%
- 23 ArmParser (Yerevan) software1-P 58.4%
- 24 ONLP lab (Ra'anana) software3-P 55.5%
- 25 HUJI (Yerushalayim) software1-P 1.89%
- 26 SParse (İstanbul) software1-P 0%
Source: Official UD Ancient Greek PROIEL leaderboard
| # | Entry | Score | Setting | Date | Source |
|---|---|---|---|---|---|
| 1 | HIT-SCIR (Harbin) | 79.3% | software1-P | Official leaderboard | |
| 2 | Stanford (Stanford) | 76.8% | software2 | Official leaderboard | |
| 3 | ICS PAS (Warszawa) | 76.5% | software1-P | Official leaderboard | |
| 4 | TurkuNLP (Turku) | 75.9% | software1-P | Official leaderboard | |
| 5 | UDPipe Future (Praha) | 75.8% | software1-P | Official leaderboard | |
| 6 | CEA LIST (Paris) | 75% | software1-P | Official leaderboard | |
| 7 | ParisNLP (Paris) | 74.5% | software1-P | Official leaderboard | |
| 8 | LATTICE (Paris) | 74.5% | software1-P | Official leaderboard | |
| 9 | AntNLP (Shanghai) | 73.8% | software1-P | Official leaderboard | |
| 10 | NLP-Cube (București) | 73% | software1-P | Official leaderboard | |
| 11 | iParse (Pittsburgh) | 72.5% | software1-P | Official leaderboard | |
| 12 | Uppsala (Uppsala) | 72.2% | software1-P | Official leaderboard | |
| 13 | IBM NY (Yorktown Heights) | 71.8% | software1-P | Official leaderboard | |
| 14 | SLT-Interactions (Bengaluru) | 71.6% | software2-P | Official leaderboard | |
| 15 | LeisureX (Shanghai) | 71% | software1-P | Official leaderboard | |
| 16 | Fudan (Shanghai) | 69.5% | software5-P | Official leaderboard | |
| 17 | BASELINE UDPipe 1.2 (Praha) | 67.6% | software1-P | Official leaderboard | |
| 18 | UniMelb (Melbourne) | 67.3% | software1-P | Official leaderboard | |
| 19 | Phoenix (Shanghai) | 66.1% | software1-P | Official leaderboard | |
| 20 | BOUN (İstanbul) | 64.8% | software2-P | Official leaderboard | |
| 21 | CUNI x-ling (Praha) | 63.3% | software1-P | Official leaderboard | |
| 22 | KParse (İstanbul) | 62.1% | software1-P | Official leaderboard | |
| 23 | ArmParser (Yerevan) | 58.4% | software1-P | Official leaderboard | |
| 24 | ONLP lab (Ra'anana) | 55.5% | software3-P | Official leaderboard | |
| 25 | HUJI (Yerushalayim) | 1.89% | software1-P | Official leaderboard | |
| 26 | SParse (İstanbul) | 0% | software1-P | Official leaderboard |
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Multilingual
- MMMLU In use MMLU professionally translated into 14 languages. Multiple-choice accuracy. Claude Mythos Preview 92.7%
- WMT24++ In use English-to-55-language translation of WMT24 test data. Scored with automatic metrics such as MetricX or xCOMET.
K-EXAONE 90.5%
- MMLU-ProX In use MMLU-Pro translated into up to 29 languages with expert review. Multiple-choice accuracy per language.
Qwen3.7-Max 87%
- CLIcK In use Models answer 1,995 multiple-choice questions on Korean language and culture. Questions come from official Korean exams and textbooks in eleven categories.
Solar Open 2 90.7%
- Global MMLU Lite 3.0 In use Models answer 400 human-translated MMLU questions per language in 23 languages. Version 3.0 adds Czech, Hungarian, Oriya, Slovak and Tajik and updates Italian. Accuracy. Claude Opus 5 95.7%
- MMMB In use Models answer multiple-choice questions about images in six languages across 15 categories. The set has about 12,000 questions. LFM2.5-VL-3B 83%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the official leaderboard (26 entries).
- Pass Models scored 26 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom No scores yet.
Leaderboard import 3 pass
- Pass Importer We read the board on universaldependencies.org: 26 entries at the last read, none a model we track.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.