MATH
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 including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.
o3-mini from OpenAI currently leads the MATH leaderboard with a score of 0.979 across 70 evaluated AI models.
What MATH measures
MATH is a text benchmark that evaluates large language models on math and reasoning tasks. LLM Stats tracks 70 models on this benchmark, with a maximum possible score of 1. Current average across reported models is 0.7, with the leader reaching 1.0.
Compare leaders on the best AI for math and best AI for reasoning leaderboards.
Publication
- Paper
- Measuring Mathematical Problem Solving With the MATH Dataset
- Authors
- Dan Hendrycks, Collin Burns, Saurav Kadavath, Akul Arora, and 4 others
- Published
- arXiv
- 2103.03874
Abstract
Many intellectual endeavors require mathematical problem solving, but this skill remains beyond the capabilities of computers. To measure this ability in machine learning models, we introduce MATH, a new dataset of 12,500 challenging competition mathematics problems. Each problem in MATH has a full step-by-step solution which can be used to teach models to generate answer derivations and explanations. To facilitate future research and increase accuracy on MATH, we also contribute a large auxiliary pretraining dataset which helps teach models the fundamentals of mathematics. Even though we are able to increase accuracy on MATH, our results show that accuracy remains relatively low, even with enormous Transformer models. Moreover, we find that simply increasing budgets and model parameter counts will be impractical for achieving strong mathematical reasoning if scaling trends continue. While scaling Transformers is automatically solving most other text-based tasks, scaling is not currently solving MATH. To have more traction on mathematical problem solving we will likely need new algorithmic advancements from the broader research community.
o3-mini leads with 97.9%, followed by o1 at 96.4% and 
MiniStral 3 (14B Instruct 2512) at 90.4%.
Progress Over Time
Interactive timeline showing model performance evolution on MATH
MATH Leaderboard
| Context | Cost | License | ||||
|---|---|---|---|---|---|---|
| 1 | OpenAI | — | — | — | ||
| 2 | OpenAI | — | — | — | ||
| 3 | Mistral AI | 14B | — | — | ||
| 3 | Mistral AI | 675B | — | — | ||
| 5 |  Google | — | — | — | ||
| 6 |  Moonshot AI | 1.0T | — | — | ||
| 7 | Google | 27B | — | — | ||
| 8 | Mistral AI | 8B | — | — | ||
| 9 | Google | — | — | — | ||
| 10 | Google | — | — | — | ||
| 11 | OpenAI | — | — | — | ||
| 12 | OpenAI | — | — | — | ||
| 13 | Google | 12B | — | — | ||
| 14 |  Alibaba Cloud / Qwen Team | 73B | — | — | ||
| 14 | Alibaba Cloud / Qwen Team | 33B | — | — | ||
| 16 | Mistral AI | 3B | — | — | ||
| 17 | Alibaba Cloud / Qwen Team | 34B | — | — | ||
| 18 |  Microsoft | 15B | — | — | ||
| 19 | Alibaba Cloud / Qwen Team | 15B | — | — | ||
| 20 |  Anthropic | — | — | — | ||
| 21 | Google | — | — | — | ||
| 22 | 70B | — | — | |||
| 23 |  Amazon | — | — | — | ||
| 23 | OpenAI | — | 128K | $2.50 / $10.00 | ||
| 25 |  xAI | — | — | — | ||
| 26 | Google | 4B | — | — | ||
| 27 | Alibaba Cloud / Qwen Team | 8B | — | — | ||
| 28 |  DeepSeek | 236B | — | — | ||
| 29 | 405B | — | — | |||
| 30 | Amazon | — | — | — | ||
| 31 | xAI | — | — | — | ||
| 32 | OpenAI | — | 128K | $10.00 / $30.00 | ||
| 33 | Alibaba Cloud / Qwen Team | 235B | — | — | ||
| 34 | Alibaba Cloud / Qwen Team | 7B | — | — | ||
| 35 | Anthropic | — | — | — | ||
| 36 | Mistral AI | 24B | — | — | ||
| 37 | OpenAI | — | — | — | ||
| 37 | Moonshot AI | 1.0T | — | — | ||
| 39 | Mistral AI | 24B | — | — | ||
| 40 | Anthropic | — | — | — | ||
| 41 | Mistral AI | 24B | — | — | ||
| 41 | Amazon | — | — | — | ||
| 43 | 90B | — | — | |||
| 44 | Microsoft | 4B | — | — | ||
| 45 |  Meta | 400B | — | — | ||
| 46 | Anthropic | — | — | — | ||
| 47 | Alibaba Cloud / Qwen Team | 72B | — | — | ||
| 48 | Microsoft | 60B | — | — | ||
| 49 | Google | 8B | — | — | ||
| 50 | Alibaba Cloud / Qwen Team | 32B | — | — |
FAQ
Common questions about MATH.
More evaluations to explore
Related benchmarks in the same category
A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.
A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks. Features over 12,000 curated questions across 14 domains and causes a 16-33% accuracy drop compared to original MMLU.
All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999. Used as an AI benchmark to evaluate large language models' ability to solve complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning.
Massive Multitask Language Understanding benchmark testing knowledge across 57 diverse subjects including STEM, humanities, social sciences, and professional domains
A verified subset of 500 software engineering problems from real GitHub issues, validated by human annotators for evaluating language models' ability to resolve real-world coding issues by generating patches for Python codebases.
Humanity's Last Exam (HLE) is a multi-modal academic benchmark with 2,500 questions across mathematics, humanities, and natural sciences, designed to test LLM capabilities at the frontier of human knowledge with unambiguous, verifiable solutions