math benchmark
MMLU-STEM
STEM-focused subset of the Massive Multitask Language Understanding benchmark, evaluating language models on science, technology, engineering, and mathematics topics including physics, chemistry, mathematics, and other technical subjects.
Updated Aug 7, 2026
Published models2
Registry coverage2
MetricScore
EvidenceB
MMLU-STEM leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | AC | 80.9% | 100.0% | 2 | C | |
| 2 | AC | 76.4% | 0.0% | 2 | C |
Score distribution
Top published rows on the benchmark's original scale.
MMLU-STEM highlights
The top published results on this benchmark's own scale.
What is MMLU-STEM?
Definition and scoring fields from the benchmark registry.
STEM-focused subset of the Massive Multitask Language Understanding benchmark, evaluating language models on science, technology, engineering, and mathematics topics including physics, chemistry, mathematics, and other technical subjects.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MMLU-STEM
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- mmlu-stem|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MMLU-STEM.
Which model scores highest on MMLU-STEM?
Qwen2.5 32B Instruct is currently ranked first with 80.9%.
What does MMLU-STEM measure?
STEM-focused subset of the Massive Multitask Language Understanding benchmark, evaluating language models on science, technology, engineering, and mathematics topics including physics, chemistry, mathematics, and other technical subjects.
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.