math benchmark
PolyMath-en
PolyMath is a multilingual mathematical reasoning benchmark covering 18 languages and 4 difficulty levels from easy to hard, ensuring difficulty comprehensiveness, language diversity, and high-quality translation. The benchmark evaluates mathematical reasoning capabilities of large language models across diverse linguistic contexts, making it a highly discriminative multilingual mathematical benchmark.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
PolyMath-en leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | MA | 65.1% | 100.0% | 2 | C | |
| 2 | MA | 65.1% | 0.0% | 2 | C |
Score distribution
Top published rows on the benchmark's original scale.
PolyMath-en highlights
The top published results on this benchmark's own scale.
What is PolyMath-en?
Definition and scoring fields from the benchmark registry.
PolyMath is a multilingual mathematical reasoning benchmark covering 18 languages and 4 difficulty levels from easy to hard, ensuring difficulty comprehensiveness, language diversity, and high-quality translation. The benchmark evaluates mathematical reasoning capabilities of large language models across diverse linguistic contexts, making it a highly discriminative multilingual mathematical benchmark.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- PolyMath-en
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- polymath-en|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about PolyMath-en.
Which model scores highest on PolyMath-en?
Kimi K2 Instruct is currently ranked first with 65.1%.
What does PolyMath-en measure?
PolyMath is a multilingual mathematical reasoning benchmark covering 18 languages and 4 difficulty levels from easy to hard, ensuring difficulty comprehensiveness, language diversity, and high-quality translation. The benchmark evaluates mathematical reasoning capabilities of large language models across diverse linguistic contexts, making it a highly discriminative multilingual mathematical benchmark.
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.