math benchmark
STEM Leaderboard
A comprehensive multimodal benchmark dataset with 448 skills and 1,073,146 questions spanning all STEM subjects (Science, Technology, Engineering, Mathematics), designed to test neural models' vision-language STEM skills based on K-12 curriculum. Unlike existing datasets that focus on expert-level ability, this dataset includes fundamental skills designed around educational standards.
Updated Aug 17, 2026
Models1
Model coverage1
MetricScore
EvidenceB
STEM Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score34.0% | Percentile100.0% | Participants1 | EvidenceC | Evaluated |
STEM Highlights
The leading models and scores on this benchmark.
The Top AI Models for STEM
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis stem AI model leaderboard uses descending score in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01ACQwen2.5-Coder 7B InstructAlibaba Cloud / Qwen Team
- Score
- 34.0%
- Price
- $0.14 input / $0.29 output per 1M tokens
Strengths
- Ranks #1 of 1 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures STEM, not total model capability
Selection summary
Best AI Models for STEM
Qwen2.5-Coder 7B Instruct currently leads STEM with 34.0%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is STEM?
What STEM measures and how its scores work.
A comprehensive multimodal benchmark dataset with 448 skills and 1,073,146 questions spanning all STEM subjects (Science, Technology, Engineering, Mathematics), designed to test neural models' vision-language STEM skills based on K-12 curriculum. Unlike existing datasets that focus on expert-level ability, this dataset includes fundamental skills designed around educational standards.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- STEM
- Modality
- multimodal
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- stem|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about STEM.
Which model scores highest on STEM?
Qwen2.5-Coder 7B Instruct is currently ranked first with 34.0%.
What does STEM measure?
A comprehensive multimodal benchmark dataset with 448 skills and 1,073,146 questions spanning all STEM subjects (Science, Technology, Engineering, Mathematics), designed to test neural models' vision-language STEM skills based on K-12 curriculum. Unlike existing datasets that focus on expert-level ability, this dataset includes fundamental skills designed around educational standards.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
1 model results are currently shown.
Does this benchmark affect the overall score?
No. This benchmark is shown for reference but does not contribute to the overall score.