language benchmark
MEGA UDPOS
Universal Dependencies POS tagging as part of the MEGA benchmark suite. A multilingual part-of-speech tagging dataset based on Universal Dependencies treebanks, utilizing the universal POS tag set of 17 tags across 38 diverse languages from different language families. Used for evaluating multilingual POS tagging systems.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
MEGA UDPOS leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MI | 60.4% | 100.0% | 2 | C | |
| 2 | MI | 46.5% | 0.0% | 2 | C |
Score distribution
Top published rows on the benchmark's original scale.
MEGA UDPOS highlights
The top published results on this benchmark's own scale.
What is MEGA UDPOS?
Definition and scoring fields from the benchmark registry.
Universal Dependencies POS tagging as part of the MEGA benchmark suite. A multilingual part-of-speech tagging dataset based on Universal Dependencies treebanks, utilizing the universal POS tag set of 17 tags across 38 diverse languages from different language families. Used for evaluating multilingual POS tagging systems.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MEGA UDPOS
- Modality
- text
- Primary category
- language
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mega-udpos|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MEGA UDPOS.
Which model scores highest on MEGA UDPOS?
Phi-3.5-MoE-instruct is currently ranked first with 60.4%.
What does MEGA UDPOS measure?
Universal Dependencies POS tagging as part of the MEGA benchmark suite. A multilingual part-of-speech tagging dataset based on Universal Dependencies treebanks, utilizing the universal POS tag set of 17 tags across 38 diverse languages from different language families. Used for evaluating multilingual POS tagging systems.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
2 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.