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 DirectoryCompare Models

Scoring & Data

Scoring & Data
393 models668 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 BenchmarksReasoningMath

Vendors

All VendorsOpenAIAnthropicGoogle
llmboard.aiCopyright 2026 llmboard.ai

legal benchmark

SuperGPQA Leaderboard

SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.

Updated Aug 17, 2026

Models34
Model coverage34
MetricScore
EvidenceB

On this page

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

SuperGPQA Ranking

Higher score ranks better on this benchmark.

30 of 34 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelACQwen3.7 MaxAlibaba Cloud / Qwen TeamScore73.6%Percentile100.0%Participants34EvidenceCEvaluatedAug 17, 2026
Rank02ModelACQwen3.6 PlusAlibaba Cloud / Qwen TeamScore71.6%Percentile97.0%Participants34EvidenceCEvaluatedAug 17, 2026
Rank03ModelACQwen3.7-PlusAlibaba Cloud / Qwen TeamScore71.4%Percentile93.9%Participants34EvidenceCEvaluatedAug 17, 2026
Rank04ModelBYSeed 2.1 ProByteDanceScore70.8%Percentile90.9%Participants34EvidenceCEvaluatedAug 17, 2026
Rank05ModelACQwen3.5-397B-A17BAlibaba Cloud / Qwen TeamScore70.4%Percentile87.9%Participants34EvidenceCEvaluatedAug 17, 2026
Rank06ModelBYSeed 2.1 TurboByteDanceScore67.4%Percentile84.8%Participants34EvidenceCEvaluatedAug 17, 2026
Rank07ModelACQwen3.5-122B-A10BAlibaba Cloud / Qwen TeamScore67.1%Percentile81.8%Participants34EvidenceCEvaluatedAug 17, 2026
Rank08ModelACQwen3.6-27BAlibaba Cloud / Qwen TeamScore66.0%Percentile78.8%Participants34EvidenceCEvaluatedAug 17, 2026
Rank09ModelACQwen3.5-27BAlibaba Cloud / Qwen TeamScore65.6%Percentile75.8%Participants34EvidenceCEvaluatedAug 17, 2026
Rank10ModelACQwen3 MaxAlibaba Cloud / Qwen TeamScore65.1%Percentile72.7%Participants34EvidenceCEvaluatedAug 17, 2026
Rank11ModelACQwen3-235B-A22B-Thinking-2507Alibaba Cloud / Qwen TeamScore64.9%Percentile69.7%Participants34EvidenceCEvaluatedAug 17, 2026
Rank12ModelACQwen3.6-35B-A3BAlibaba Cloud / Qwen TeamScore64.7%Percentile66.7%Participants34EvidenceCEvaluatedAug 17, 2026
Rank13ModelACQwen3 VL 235B A22B ThinkingAlibaba Cloud / Qwen TeamScore64.3%Percentile63.6%Participants34EvidenceCEvaluatedAug 17, 2026
Rank14ModelACQwen3.5-35B-A3BAlibaba Cloud / Qwen TeamScore63.4%Percentile60.6%Participants34EvidenceCEvaluatedAug 17, 2026
Rank15ModelACQwen3-235B-A22B-Instruct-2507Alibaba Cloud / Qwen TeamScore62.6%Percentile57.6%Participants34EvidenceCEvaluatedAug 17, 2026
Rank16ModelACQwen3-Next-80B-A3B-ThinkingAlibaba Cloud / Qwen TeamScore60.8%Percentile54.5%Participants34EvidenceCEvaluatedAug 17, 2026
Rank17ModelACQwen3 VL 235B A22B InstructAlibaba Cloud / Qwen TeamScore60.4%Percentile51.5%Participants34EvidenceCEvaluatedAug 17, 2026
Rank18ModelACQwen3 VL 32B ThinkingAlibaba Cloud / Qwen TeamScore59.0%Percentile48.5%Participants34EvidenceCEvaluatedAug 17, 2026
Rank19ModelACQwen3-Next-80B-A3B-InstructAlibaba Cloud / Qwen TeamScore58.8%Percentile45.5%Participants34EvidenceCEvaluatedAug 17, 2026
Rank20ModelACQwen3.5-9BAlibaba Cloud / Qwen TeamScore58.2%Percentile42.4%Participants34EvidenceCEvaluatedAug 17, 2026
Rank21ModelMAKimi K2 InstructMoonshot AIScore57.2%Percentile39.4%Participants34EvidenceCEvaluatedAug 17, 2026
Rank22ModelMAKimi K2-Instruct-0905Moonshot AIScore57.2%Percentile36.4%Participants34EvidenceCEvaluatedAug 17, 2026
Rank23ModelACQwen3 VL 30B A3B ThinkingAlibaba Cloud / Qwen TeamScore56.4%Percentile33.3%Participants34EvidenceCEvaluatedAug 17, 2026
Rank24ModelACQwen3 VL 32B InstructAlibaba Cloud / Qwen TeamScore54.6%Percentile30.3%Participants34EvidenceCEvaluatedAug 17, 2026
Rank25ModelACQwen3 VL 30B A3B InstructAlibaba Cloud / Qwen TeamScore53.1%Percentile27.3%Participants34EvidenceCEvaluatedAug 17, 2026
Rank26ModelACQwen3.5-4BAlibaba Cloud / Qwen TeamScore52.9%Percentile24.2%Participants34EvidenceCEvaluatedAug 17, 2026
Rank27ModelACQwen3 VL 8B ThinkingAlibaba Cloud / Qwen TeamScore51.2%Percentile21.2%Participants34EvidenceCEvaluatedAug 17, 2026
Rank28ModelACQwen3 VL 4B ThinkingAlibaba Cloud / Qwen TeamScore46.8%Percentile18.2%Participants34EvidenceCEvaluatedAug 17, 2026
Rank29ModelMAKimi K2 BaseMoonshot AIScore44.7%Percentile15.2%Participants34EvidenceCEvaluatedAug 17, 2026
Rank30ModelACQwen3 VL 8B InstructAlibaba Cloud / Qwen TeamScore44.5%Percentile12.1%Participants34EvidenceCEvaluatedAug 17, 2026
Rank31ModelACQwen3 235B A22BAlibaba Cloud / Qwen TeamScore44.1%Percentile9.1%Participants34EvidenceCEvaluatedAug 17, 2026
Rank32ModelACQwen3 VL 4B InstructAlibaba Cloud / Qwen TeamScore40.3%Percentile6.1%Participants34EvidenceCEvaluatedAug 17, 2026
Rank33ModelACQwen3.5-2BAlibaba Cloud / Qwen TeamScore37.5%Percentile3.0%Participants34EvidenceCEvaluatedAug 17, 2026
Rank34ModelACQwen3.5-0.8BAlibaba Cloud / Qwen TeamScore21.3%Percentile0.0%Participants34EvidenceCEvaluatedAug 17, 2026

SuperGPQA Highlights

The leading models and scores on this benchmark.

Rank #1Qwen3.7 Max73.6%Rank #2Qwen3.6 Plus71.6%Rank #3Qwen3.7-Plus71.4%Rank #4Seed 2.1 Pro70.8%

SuperGPQA Score Distribution

A closer view of the leading scores on this benchmark.

SuperGPQA

The Top AI Models for SuperGPQA

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

Ranking basisThis supergpqa 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.

  1. 01
    AC
    Qwen3.7 MaxAlibaba Cloud / Qwen Team
    Score
    73.6%
    Price
    $2.5 input / $7.5 output per 1M tokens
    Speed
    Up to 5.8 tok/s via Together

    Strengths

    • Ranks #1 of 34 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SuperGPQA, not total model capability
  2. 02
    AC
    Qwen3.6 PlusAlibaba Cloud / Qwen Team
    Score
    71.6%
    Price
    $0.50 input / $3.0 output per 1M tokens
    Speed
    Up to 16 tok/s via Together

    Strengths

    • Ranks #2 of 34 compared models
    • 97th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SuperGPQA, not total model capability
  3. 03
    AC
    Qwen3.7-PlusAlibaba Cloud / Qwen Team
    Score
    71.4%
    Price
    $0.50 input / $3.0 output per 1M tokens

    Strengths

    • Ranks #3 of 34 compared models
    • 94th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SuperGPQA, not total model capability
  4. 04
    BY
    Seed 2.1 ProByteDance
    Score
    70.8%

    Strengths

    • Ranks #4 of 34 compared models
    • 91th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SuperGPQA, not total model capability
  5. 05
    AC
    Qwen3.5-397B-A17BAlibaba Cloud / Qwen Team
    Score
    70.4%
    Price
    $0.60 input / $3.6 output per 1M tokens

    Strengths

    • Ranks #5 of 34 compared models
    • 88th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SuperGPQA, not total model capability

Selection summary

Best AI Models for SuperGPQA

Qwen3.7 Max currently leads SuperGPQA with 73.6%. 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.

Benchmark rank #1Qwen3.7 Max73.6% · $2.5 input / $7.5 output per 1M tokensBenchmark rank #2Qwen3.6 Plus71.6% · $0.50 input / $3.0 output per 1M tokensBenchmark rank #3Qwen3.7-Plus71.4% · $0.50 input / $3.0 output per 1M tokens

What is SuperGPQA?

What SuperGPQA measures and how its scores work.

SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.

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

Family
SuperGPQA
Modality
text
Primary category
legal
Score direction
higher
LLMBoard eligible
Yes
Evaluation key
supergpqa|llm-stats-current

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

FAQ

Common questions about SuperGPQA.

Which model scores highest on SuperGPQA?

Qwen3.7 Max is currently ranked first with 73.6%.

What does SuperGPQA measure?

SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines. The benchmark contains 25,957 questions covering 13 broad disciplinary areas including Engineering, Medicine, Science, and Law, with specialized fields in light industry, agriculture, and service-oriented domains. It employs a Human-LLM collaborative filtering mechanism with over 80 expert annotators to create challenging questions that assess graduate-level knowledge and reasoning capabilities.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

34 model results are currently shown.

Does this benchmark affect the overall score?

Yes. This benchmark can contribute to the current LLMBoard capability score.