math benchmark
MATH (CoT)
MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects. This variant uses Chain-of-Thought prompting to encourage step-by-step reasoning.
Updated Aug 7, 2026
Published models6
Registry coverage6
MetricScore
EvidenceB
MATH (CoT) leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | ME | 68.0% | 100.0% | 6 | C | |
| 2 | MA | 67.6% | 80.0% | 6 | C | |
| 3 | MA | 67.6% | 60.0% | 6 | C | |
| 4 | MA | 62.6% | 40.0% | 6 | C | |
| 5 | MA | 60.1% | 20.0% | 6 | C | |
| 6 | ME | 51.9% | 0.0% | 6 | C |
Score distribution
Top published rows on the benchmark's original scale.
MATH (CoT) highlights
The top published results on this benchmark's own scale.
What is MATH (CoT)?
Definition and scoring fields from the benchmark registry.
MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects. This variant uses Chain-of-Thought prompting to encourage step-by-step reasoning.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- MATH (CoT)
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- math-(cot)|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about MATH (CoT).
Which model scores highest on MATH (CoT)?
Llama 3.1 70B Instruct is currently ranked first with 68.0%.
What does MATH (CoT) measure?
MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions. Each problem includes full step-by-step solutions and spans multiple difficulty levels (1-5) across seven mathematical subjects. This variant uses Chain-of-Thought prompting to encourage step-by-step reasoning.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
6 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.