math benchmark
MATH Leaderboard
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.
Updated Aug 17, 2026
Models71
Model coverage71
MetricScore
EvidenceB
MATH Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelOP | Score97.9% | Percentile100.0% | Participants71 | EvidenceC | Evaluated |
| Rank02 | ModelOP | Score96.4% | Percentile98.6% | Participants71 | EvidenceC | Evaluated |
| Rank03 | ModelMA | Score90.4% | Percentile97.1% | Participants71 | EvidenceC | Evaluated |
| Rank04 | ModelMA | Score90.4% | Percentile95.7% | Participants71 | EvidenceC | Evaluated |
| Rank05 | ModelGO | Score89.7% | Percentile94.3% | Participants71 | EvidenceC | Evaluated |
| Rank06 | ModelMA | Score89.1% | Percentile92.9% | Participants71 | EvidenceC | Evaluated |
| Rank07 | ModelGO | Score89.0% | Percentile91.4% | Participants71 | EvidenceC | Evaluated |
| Rank08 | ModelMA | Score87.6% | Percentile90.0% | Participants71 | EvidenceC | Evaluated |
| Rank09 | ModelGO | Score86.8% | Percentile88.6% | Participants71 | EvidenceC | Evaluated |
| Rank10 | ModelGO | Score86.5% | Percentile87.1% | Participants71 | EvidenceC | Evaluated |
| Rank11 | ModelXI | Score86.2% | Percentile85.7% | Participants71 | EvidenceC | Evaluated |
| Rank12 | ModelOP | Score85.5% | Percentile84.3% | Participants71 | EvidenceC | Evaluated |
| Rank13 | ModelOP | Score84.7% | Percentile82.9% | Participants71 | EvidenceC | Evaluated |
| Rank14 | ModelGO | Score83.8% | Percentile81.4% | Participants71 | EvidenceC | Evaluated |
| Rank15 | ModelAC | Score83.1% | Percentile80.0% | Participants71 | EvidenceC | Evaluated |
| Rank16 | ModelAC | Score83.1% | Percentile78.6% | Participants71 | EvidenceC | Evaluated |
| Rank17 | ModelMA | Score83.0% | Percentile77.1% | Participants71 | EvidenceC | Evaluated |
| Rank18 | ModelAC | Score82.2% | Percentile75.7% | Participants71 | EvidenceC | Evaluated |
| Rank19 | ModelMI | Score80.4% | Percentile74.3% | Participants71 | EvidenceC | Evaluated |
| Rank20 | ModelAC | Score80.0% | Percentile72.9% | Participants71 | EvidenceC | Evaluated |
| Rank21 | ModelAN | Score78.3% | Percentile71.4% | Participants71 | EvidenceC | Evaluated |
| Rank22 | ModelGO | Score77.9% | Percentile70.0% | Participants71 | EvidenceC | Evaluated |
| Rank23 | ModelME | Score77.0% | Percentile68.6% | Participants71 | EvidenceC | Evaluated |
| Rank24 | ModelOP | Score76.6% | Percentile67.1% | Participants71 | EvidenceC | Evaluated |
| Rank25 | ModelAM Nova ProAmazon | Score76.6% | Percentile65.7% | Participants71 | EvidenceC | Evaluated |
| Rank26 | ModelXA | Score76.1% | Percentile64.3% | Participants71 | EvidenceC | Evaluated |
| Rank27 | ModelGO | Score75.6% | Percentile62.9% | Participants71 | EvidenceC | Evaluated |
| Rank28 | ModelAC | Score75.5% | Percentile61.4% | Participants71 | EvidenceC | Evaluated |
| Rank29 | ModelDE | Score74.7% | Percentile60.0% | Participants71 | EvidenceC | Evaluated |
| Rank30 | ModelME | Score73.8% | Percentile58.6% | Participants71 | EvidenceC | Evaluated |
| Rank31 | ModelAMNova LiteAmazon | Score73.3% | Percentile57.1% | Participants71 | EvidenceC | Evaluated |
| Rank32 | ModelXA | Score73.0% | Percentile55.7% | Participants71 | EvidenceC | Evaluated |
| Rank33 | ModelOP | Score72.6% | Percentile54.3% | Participants71 | EvidenceC | Evaluated |
| Rank34 | ModelAC | Score71.8% | Percentile52.9% | Participants71 | EvidenceC | Evaluated |
| Rank35 | ModelAC | Score71.5% | Percentile51.4% | Participants71 | EvidenceC | Evaluated |
| Rank36 | ModelAN | Score71.1% | Percentile50.0% | Participants71 | EvidenceC | Evaluated |
| Rank37 | ModelMA | Score70.6% | Percentile48.6% | Participants71 | EvidenceC | Evaluated |
| Rank38 | ModelOP | Score70.2% | Percentile47.1% | Participants71 | EvidenceC | Evaluated |
| Rank39 | ModelMA | Score70.2% | Percentile45.7% | Participants71 | EvidenceC | Evaluated |
| Rank40 | ModelMA | Score69.4% | Percentile44.3% | Participants71 | EvidenceC | Evaluated |
| Rank41 | ModelAN | Score69.4% | Percentile42.9% | Participants71 | EvidenceC | Evaluated |
| Rank42 | ModelMA | Score69.3% | Percentile41.4% | Participants71 | EvidenceC | Evaluated |
| Rank43 | ModelAMNova MicroAmazon | Score69.3% | Percentile40.0% | Participants71 | EvidenceC | Evaluated |
| Rank44 | ModelME | Score68.0% | Percentile38.6% | Participants71 | EvidenceC | Evaluated |
| Rank45 | ModelMI | Score64.0% | Percentile37.1% | Participants71 | EvidenceC | Evaluated |
| Rank46 | ModelME | Score61.2% | Percentile35.7% | Participants71 | EvidenceC | Evaluated |
| Rank47 | ModelAN | Score60.1% | Percentile34.3% | Participants71 | EvidenceC | Evaluated |
| Rank48 | ModelAC | Score59.7% | Percentile32.9% | Participants71 | EvidenceC | Evaluated |
| Rank49 | ModelMI | Score59.5% | Percentile31.4% | Participants71 | EvidenceC | Evaluated |
| Rank50 | ModelGO | Score58.7% | Percentile30.0% | Participants71 | EvidenceC | Evaluated |
| Rank51 | ModelAC | Score57.2% | Percentile28.6% | Participants71 | EvidenceC | Evaluated |
| Rank52 | ModelMA | Score54.5% | Percentile27.1% | Participants71 | EvidenceC | Evaluated |
| Rank53 | ModelME | Score51.9% | Percentile25.7% | Participants71 | EvidenceC | Evaluated |
| Rank54 | ModelXA | Score50.6% | Percentile24.3% | Participants71 | EvidenceC | Evaluated |
| Rank55 | ModelME | Score50.3% | Percentile22.9% | Participants71 | EvidenceC | Evaluated |
| Rank56 | ModelAC | Score49.6% | Percentile21.4% | Participants71 | EvidenceC | Evaluated |
| Rank57 | ModelMI | Score48.5% | Percentile20.0% | Participants71 | EvidenceC | Evaluated |
| Rank58 | ModelMA | Score48.1% | Percentile18.6% | Participants71 | EvidenceC | Evaluated |
| Rank59 | ModelGO | Score48.0% | Percentile17.1% | Participants71 | EvidenceC | Evaluated |
| Rank60 | ModelME | Score48.0% | Percentile15.7% | Participants71 | EvidenceC | Evaluated |
| Rank61 | ModelAC | Score46.6% | Percentile14.3% | Participants71 | EvidenceC | Evaluated |
| Rank62 | ModelMA | Score46.0% | Percentile12.9% | Participants71 | EvidenceC | Evaluated |
| Rank63 | ModelAN | Score43.1% | Percentile11.4% | Participants71 | EvidenceC | Evaluated |
| Rank64 | ModelOP | Score43.1% | Percentile10.0% | Participants71 | EvidenceB | Evaluated |
| Rank65 | ModelGO | Score42.3% | Percentile8.6% | Participants71 | EvidenceC | Evaluated |
| Rank66 | ModelOP | Score42.0% | Percentile7.1% | Participants71 | EvidenceC | Evaluated |
| Rank67 | ModelAN | Score38.9% | Percentile5.7% | Participants71 | EvidenceC | Evaluated |
| Rank68 | ModelGO | Score36.6% | Percentile4.3% | Participants71 | EvidenceC | Evaluated |
| Rank69 | ModelGO | Score32.6% | Percentile2.9% | Participants71 | EvidenceB | Evaluated |
| Rank70 | ModelNR | Score20.8% | Percentile1.4% | Participants71 | EvidenceC | Evaluated |
| Rank71 | ModelBA | Score12.4% | Percentile0.0% | Participants71 | EvidenceC | Evaluated |
MATH Highlights
The leading models and scores on this benchmark.
MATH Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for MATH
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis math 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.
- 01OPo3-miniOpenAI
- Score
- 97.9%
- Price
- $1.1 input / $4.4 output per 1M tokens
- Speed
- Up to 115 tok/s via Azure
Strengths
- Ranks #1 of 71 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures MATH, not total model capability
- 02OPo1OpenAI
- Score
- 96.4%
- Price
- $15 input / $60 output per 1M tokens
- Speed
- Up to 66 tok/s via OpenAI
Strengths
- Ranks #2 of 71 compared models
- 99th percentile on this benchmark
- C evidence result
Considerations
- This result measures MATH, not total model capability
- 03MAMiniStral 3 (14B Instruct 2512)Mistral AI
- Score
- 90.4%
- Price
- $0.10 input / $0.10 output per 1M tokens
- Speed
- Up to 238 tok/s via Mistral AI
Strengths
- Ranks #3 of 71 compared models
- 97th percentile on this benchmark
- C evidence result
Considerations
- This result measures MATH, not total model capability
- 04MAMistral Large 3Mistral AI
- Score
- 90.4%
- Price
- $0.50 input / $1.5 output per 1M tokens
- Speed
- Up to 0.78 tok/s via Mistral AI
Strengths
- Ranks #4 of 71 compared models
- 96th percentile on this benchmark
- C evidence result
Considerations
- This result measures MATH, not total model capability
- 05GOGemini 2.0 FlashGoogle
- Score
- 89.7%
- Speed
- Up to 183 tok/s via Google
Strengths
- Ranks #5 of 71 compared models
- 94th percentile on this benchmark
- C evidence result
Considerations
- This result measures MATH, not total model capability
Selection summary
Best AI Models for MATH
o3-mini currently leads MATH with 97.9%. 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 MATH?
What MATH measures and how its scores work.
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.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- MATH
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- math|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about MATH.
Which model scores highest on MATH?
o3-mini is currently ranked first with 97.9%.
What does MATH 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 including Prealgebra, Algebra, Number Theory, Counting and Probability, Geometry, Intermediate Algebra, and Precalculus.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
71 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.