llmboard.aiAI model intelligence
Home

Model Rankings

OverallOpen ModelsAgentCodingReasoningMathKnowledgeInstruction FollowingTextVision
Image GenerationImage Editing
Video GenerationImage to VideoVideo Editing
Text to SpeechSpeech to Text
Embeddings

Efficiency

Chat Token PricingImage PricingVideo PricingAudio Pricing
Chat Speed & LatencyProvider Reliability

Benchmarks

GPQAMMLU-ProAIME 2025SWE-Bench VerifiedMMLUHumanity's Last ExamLiveCodeBenchMATHHumanEvalMMMU-Pro
All Benchmarks

Tools

Model Directory

Scoring & Data

Scoring & Data
1224 models729 benchmarks

Leaderboard Center

Overall RankingCodingCore BenchmarksPrice & ValueRuntime Performance

Modalities

All ModelsImage GenerationImage EditingVideo GenerationImage-to-VideoVideo EditingText-to-SpeechSpeech-to-TextEmbeddings

Data & Methods

Scoring MethodAll Benchmarks
llmboard.aiCopyright 2026 llmboard.ai

math benchmark

AIME 2025 Leaderboard

AIME 2025 is an AI benchmark based on all 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II).

Updated Sep 7, 2026

Models100
Model coverage119
MetricScore
EvidenceC

On this page

  • Ranking
  • Highlights
  • Distribution
  • Top models
  • About
  • FAQ

AIME 2025 Ranking

Higher score ranks better on this benchmark.

30 of 100 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelGOGemini 3 ProGoogleScore100.00%Percentile100.00%Participants119EvidenceCEvaluatedSep 8, 2026
Rank02ModelOPGPT-5.2OpenAIScore100.00%Percentile99.15%Participants119EvidenceCEvaluatedSep 8, 2026
Rank03ModelOPGPT-5.2 ProOpenAIScore100.00%Percentile98.31%Participants119EvidenceCEvaluatedSep 8, 2026
Rank04ModelXAGrok-4 HeavyxAIScore100.00%Percentile97.46%Participants119EvidenceCEvaluatedSep 8, 2026
Rank05ModelMAKimi K2-Thinking-0905Moonshot AIScore100.00%Percentile96.61%Participants119EvidenceCEvaluatedSep 8, 2026
Rank06ModelANClaude Opus 4.6AnthropicScore99.79%Percentile95.76%Participants119EvidenceCEvaluatedSep 8, 2026
Rank07ModelGOGemini 3 FlashGoogleScore99.70%Percentile94.92%Participants119EvidenceCEvaluatedSep 8, 2026
Rank08ModelOPGPT-5.1 HighOpenAIScore99.60%Percentile94.07%Participants119EvidenceCEvaluatedSep 8, 2026
Rank09ModelMELongCat-Flash-Thinking-2601MeituanScore99.60%Percentile93.22%Participants119EvidenceCEvaluatedSep 8, 2026
Rank10ModelNVNemotron 3 Nano (30B A3B)NVIDIAScore99.20%Percentile92.37%Participants119EvidenceCEvaluatedSep 8, 2026
Rank11ModelOPGPT OSS 20B HighOpenAIScore98.70%Percentile91.53%Participants119EvidenceCEvaluatedSep 8, 2026
Rank12ModelOPGPT-5.1 MediumOpenAIScore98.40%Percentile90.68%Participants119EvidenceCEvaluatedSep 8, 2026
Rank13ModelBYSeed 2.0 ProByteDanceScore98.30%Percentile89.83%Participants119EvidenceCEvaluatedSep 8, 2026
Rank14ModelSTStep-3.5-FlashStepFunScore97.30%Percentile88.98%Participants119EvidenceCEvaluatedSep 8, 2026
Rank15ModelMIMAI-Thinking-1MicrosoftScore97.00%Percentile88.14%Participants119EvidenceCEvaluatedSep 8, 2026
Rank16ModelOPGPT-5.1 Codex HighOpenAIScore96.70%Percentile87.29%Participants119EvidenceCEvaluatedSep 8, 2026
Rank17ModelSASarvam-105BSarvam AIScore96.70%Percentile86.44%Participants119EvidenceCEvaluatedSep 8, 2026
Rank18ModelSASarvam-30BSarvam AIScore96.70%Percentile85.59%Participants119EvidenceCEvaluatedSep 8, 2026
Rank19ModelMAKimi K2.5Moonshot AIScore96.10%Percentile84.75%Participants119EvidenceCEvaluatedSep 8, 2026
Rank20ModelDEDeepSeek-V3.2-SpecialeDeepSeekScore96.00%Percentile83.90%Participants119EvidenceCEvaluatedSep 8, 2026
Rank21ModelZAGLM-4.7Zhipu AIScore95.70%Percentile83.05%Participants119EvidenceCEvaluatedSep 8, 2026
Rank22ModelOPGPT-5OpenAIScore94.60%Percentile82.20%Participants119EvidenceCEvaluatedSep 8, 2026
Rank23ModelOPGPT-5 HighOpenAIScore94.60%Percentile81.36%Participants119EvidenceCEvaluatedSep 8, 2026
Rank24ModelXIMiMo-V2-FlashXiaomiScore94.10%Percentile80.51%Participants119EvidenceCEvaluatedSep 8, 2026
Rank25ModelOPGPT-5.1OpenAIScore94.00%Percentile79.66%Participants119EvidenceCEvaluatedSep 8, 2026
Rank26ModelOPGPT-5.1 InstantOpenAIScore94.00%Percentile78.81%Participants119EvidenceCEvaluatedSep 8, 2026
Rank27ModelOPGPT-5.1 ThinkingOpenAIScore94.00%Percentile77.97%Participants119EvidenceCEvaluatedSep 8, 2026
Rank28ModelZAGLM-4.6Zhipu AIScore93.90%Percentile77.12%Participants119EvidenceCEvaluatedSep 8, 2026
Rank29ModelXAGrok-3xAIScore93.30%Percentile76.27%Participants119EvidenceCEvaluatedSep 8, 2026
Rank30ModelDEDeepSeek-V3.2 (Thinking)DeepSeekScore93.10%Percentile75.42%Participants119EvidenceCEvaluatedSep 8, 2026

AIME 2025 Highlights

The leading models and scores on this benchmark.

AIME 2025 Score Distribution

A closer view of the leading scores on this benchmark.

AIME 2025

The Top AI Models for AIME 2025

The first five results on this benchmark, with official price and output speed added where the model identity can be matched.

What is AIME 2025?

What AIME 2025 measures and how its scores work.

AIME 2025 consists of 30 olympiad-level mathematical problems from AIME I and AIME II, with integer answers from 000-999.

It measures performance on complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning, reported using the Score metric as a ratio.

Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of C.

Family
AIME 2025
Modality
text
Primary category
math
Score direction
higher
LLMBoard eligible
No
Evaluation key
overall

LLM Stats. Benchmark scores retain their original unit. Overall score eligibility is shown separately.

FAQ

Common questions about AIME 2025.

Which model scores highest on AIME 2025?

Gemini 3 Pro is currently ranked first with 100.00%.

What are the top three models on AIME 2025?

The current leaders are Gemini 3 Pro (100.00%), GPT-5.2 (100.00%), and GPT-5.2 Pro (100.00%).

Which AIME 2025 model has the lowest official input price?

Min istral 3 (3B Reasoning 2512) has the lowest matched official input price at $0.04 input / $0.04 output per 1M tokens.

Which models are fastest among AIME 2025 results?

The fastest matched records are Min istral 3 (3B Reasoning 2512) (1,222.00 tok/s via Mistral AI), GPT-5 nano (500.00 tok/s via OpenAI), and GPT-5.1 Medium (369.03 tok/s via OpenAI).

Does the highest score result prove overall model quality?

No. This benchmark measures one defined capability or task. The overall LLMBoard score uses a separate aggregation across eligible benchmark evidence.

What does AIME 2025 measure?

It measures performance on complex mathematical problems requiring multi-step logical deductions and structured symbolic reasoning, reported using the Score metric as a ratio.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

100 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.

Rank #1Gemini 3 Pro100.00%
Rank #2GPT-5.2100.00%
Rank #3GPT-5.2 Pro100.00%
Rank #4Grok-4 Heavy100.00%

Ranking basisThis aime 2025 AI model leaderboard uses descending score in the benchmark original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.

  1. 01
    GO
    Gemini 3 ProGoogle
    Score
    100.00%
    Speed
    Up to 90.00 tok/s via Google

    Strengths

    • Ranks #1 of 119 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures AIME 2025, not total model capability
  2. 02
    OP
    GPT-5.2OpenAI
    Score
    100.00%
    Price
    $1.8 input / $14 output per 1M tokens
    Speed
    Up to 100.00 tok/s via OpenAI

    Strengths

    • Ranks #2 of 119 compared models
    • 99th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures AIME 2025, not total model capability
  3. 03
    OP
    OpenAI
    Score
    100.00%
    Price
    $21 input / $168 output per 1M tokens
    Speed
    Up to 25.00 tok/s via OpenAI

    Strengths

    • Ranks #3 of 119 compared models
    • 98th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures AIME 2025, not total model capability
  4. 04
    XA
    xAI
    Score
    100.00%

    Strengths

    • Ranks #4 of 119 compared models
    • 97th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures AIME 2025, not total model capability
  5. 05
    MA
    Moonshot AI
    Score
    100.00%

    Strengths

    • Ranks #5 of 119 compared models
    • 97th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures AIME 2025, not total model capability

Selection summary

Best AI Models for AIME 2025

Gemini 3 Pro currently leads AIME 2025 with 100.00%. 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.

GPT-5.2 Pro
Grok-4 Heavy
Kimi K2-Thinking-0905
Benchmark rank #1Gemini 3 Pro100.00% · Up to 90.00 tok/s via Google
Benchmark rank #2GPT-5.2100.00% · $1.8 input / $14 output per 1M tokens
Benchmark rank #3GPT-5.2 Pro100.00% · $21 input / $168 output per 1M tokens