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

reasoning benchmark

SimpleQA Leaderboard

SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.

Updated Aug 17, 2026

Models47
Model coverage47
MetricScore
EvidenceB

On this page

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

SimpleQA Ranking

Higher score ranks better on this benchmark.

30 of 47 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelDEDeepSeek-V3.2-ExpDeepSeekScore97.1%Percentile100.0%Participants47EvidenceCEvaluatedAug 17, 2026
Rank02ModelXAGrok 4 FastxAIScore95.0%Percentile97.8%Participants47EvidenceCEvaluatedAug 17, 2026
Rank03ModelDEDeepSeek-V3.1DeepSeekScore93.4%Percentile95.7%Participants47EvidenceCEvaluatedAug 17, 2026
Rank04ModelDEDeepSeek-R1-0528DeepSeekScore92.3%Percentile93.5%Participants47EvidenceCEvaluatedAug 17, 2026
Rank05ModelBAERNIE 5.0BaiduScore75.0%Percentile91.3%Participants47EvidenceCEvaluatedAug 17, 2026
Rank06ModelGOGemini 3 ProGoogleScore72.1%Percentile89.1%Participants47EvidenceCEvaluatedAug 17, 2026
Rank07ModelGOGemini 3 FlashGoogleScore68.7%Percentile87.0%Participants47EvidenceCEvaluatedAug 17, 2026
Rank08ModelOPGPT-4.5OpenAIScore62.5%Percentile84.8%Participants47EvidenceCEvaluatedAug 17, 2026
Rank09ModelDEDeepSeek-V4-Pro-MaxDeepSeekScore57.9%Percentile82.6%Participants47EvidenceCEvaluatedAug 17, 2026
Rank10ModelACQwen3 VL 32B ThinkingAlibaba Cloud / Qwen TeamScore55.4%Percentile80.4%Participants47EvidenceCEvaluatedAug 17, 2026
Rank11ModelACQwen3-235B-A22B-Instruct-2507Alibaba Cloud / Qwen TeamScore54.3%Percentile78.3%Participants47EvidenceCEvaluatedAug 17, 2026
Rank12ModelGOGemini 2.5 Pro Preview 06-05GoogleScore54.0%Percentile76.1%Participants47EvidenceCEvaluatedAug 17, 2026
Rank13ModelACQwen3 VL 235B A22B InstructAlibaba Cloud / Qwen TeamScore51.9%Percentile73.9%Participants47EvidenceCEvaluatedAug 17, 2026
Rank14ModelGOGemini 2.5 ProGoogleScore50.8%Percentile71.7%Participants47EvidenceCEvaluatedAug 17, 2026
Rank15ModelACQwen3 VL 8B ThinkingAlibaba Cloud / Qwen TeamScore49.6%Percentile69.6%Participants47EvidenceCEvaluatedAug 17, 2026
Rank16ModelACQwen3 VL 4B InstructAlibaba Cloud / Qwen TeamScore48.0%Percentile67.4%Participants47EvidenceCEvaluatedAug 17, 2026
Rank17ModelOPo1OpenAIScore47.0%Percentile65.2%Participants47EvidenceCEvaluatedAug 17, 2026
Rank18ModelACQwen3 VL 235B A22B ThinkingAlibaba Cloud / Qwen TeamScore44.4%Percentile63.0%Participants47EvidenceCEvaluatedAug 17, 2026
Rank19ModelGOGemini 3.1 Flash-LiteGoogleScore43.3%Percentile60.9%Participants47EvidenceCEvaluatedAug 17, 2026
Rank20ModelOPo1-previewOpenAIScore42.4%Percentile58.7%Participants47EvidenceCEvaluatedAug 17, 2026
Rank21ModelOPGPT-4oOpenAIScore38.2%Percentile56.5%Participants47EvidenceCEvaluatedAug 17, 2026
Rank22ModelMAKimi K2 BaseMoonshot AIScore35.3%Percentile54.4%Participants47EvidenceCEvaluatedAug 17, 2026
Rank23ModelDEDeepSeek-V4-Flash-MaxDeepSeekScore34.1%Percentile52.2%Participants47EvidenceCEvaluatedAug 17, 2026
Rank24ModelMAKimi K2 InstructMoonshot AIScore31.0%Percentile50.0%Participants47EvidenceCEvaluatedAug 17, 2026
Rank25ModelMAKimi K2-Instruct-0905Moonshot AIScore31.0%Percentile47.8%Participants47EvidenceCEvaluatedAug 17, 2026
Rank26ModelDEDeepSeek-V4-Flash-0423DeepSeekScore28.9%Percentile45.6%Participants47EvidenceCEvaluatedAug 17, 2026
Rank27ModelACQwen3 VL 30B A3B InstructAlibaba Cloud / Qwen TeamScore27.0%Percentile43.5%Participants47EvidenceCEvaluatedAug 17, 2026
Rank28ModelGOGemini 2.5 FlashGoogleScore26.9%Percentile41.3%Participants47EvidenceCEvaluatedAug 17, 2026
Rank29ModelDEDeepSeek-V3DeepSeekScore24.9%Percentile39.1%Participants47EvidenceCEvaluatedAug 17, 2026
Rank30ModelACQwen3 VL 30B A3B ThinkingAlibaba Cloud / Qwen TeamScore23.9%Percentile37.0%Participants47EvidenceCEvaluatedAug 17, 2026
Rank31ModelMAMistral Large 3 (675B Instruct 2512 Eagle)Mistral AIScore23.8%Percentile34.8%Participants47EvidenceCEvaluatedAug 17, 2026
Rank32ModelMAMistral Large 3 (675B Base)Mistral AIScore23.8%Percentile32.6%Participants47EvidenceCEvaluatedAug 17, 2026
Rank33ModelMAMistral Large 3 (675B Instruct 2512 NVFP4)Mistral AIScore23.8%Percentile30.4%Participants47EvidenceCEvaluatedAug 17, 2026
Rank34ModelMAMistral Large 3 (675B Instruct 2512)Mistral AIScore23.8%Percentile28.3%Participants47EvidenceCEvaluatedAug 17, 2026
Rank35ModelGOGemini 2.0 Flash-LiteGoogleScore21.7%Percentile26.1%Participants47EvidenceCEvaluatedAug 17, 2026
Rank36ModelMIMiniMax M1 80KMiniMaxScore18.5%Percentile23.9%Participants47EvidenceCEvaluatedAug 17, 2026
Rank37ModelMIMiniMax M1 40KMiniMaxScore17.9%Percentile21.7%Participants47EvidenceCEvaluatedAug 17, 2026
Rank38ModelOPo3-miniOpenAIScore15.0%Percentile19.6%Participants47EvidenceCEvaluatedAug 17, 2026
Rank39ModelMAMistral Small 3.2 24B InstructMistral AIScore12.1%Percentile17.4%Participants47EvidenceCEvaluatedAug 17, 2026
Rank40ModelGOGemini 2.5 Flash-LiteGoogleScore10.7%Percentile15.2%Participants47EvidenceCEvaluatedAug 17, 2026
Rank41ModelMAMistral Small 3.1 24B InstructMistral AIScore10.4%Percentile13.0%Participants47EvidenceCEvaluatedAug 17, 2026
Rank42ModelGOGemma 3 27BGoogleScore10.0%Percentile10.9%Participants47EvidenceCEvaluatedAug 17, 2026
Rank43ModelGOGemma 3 12BGoogleScore6.3%Percentile8.7%Participants47EvidenceCEvaluatedAug 17, 2026
Rank44ModelGOGemma 3 4BGoogleScore4.0%Percentile6.5%Participants47EvidenceCEvaluatedAug 17, 2026
Rank45ModelMIPhi 4MicrosoftScore3.0%Percentile4.3%Participants47EvidenceCEvaluatedAug 17, 2026
Rank46ModelGOGemma 3 1BGoogleScore2.2%Percentile2.2%Participants47EvidenceCEvaluatedAug 17, 2026
Rank47ModelBAERNIE 4.5BaiduScore1.8%Percentile0.0%Participants47EvidenceCEvaluatedAug 17, 2026

SimpleQA Highlights

The leading models and scores on this benchmark.

Rank #1DeepSeek-V3.2-Exp97.1%Rank #2Grok 4 Fast95.0%Rank #3DeepSeek-V3.193.4%Rank #4DeepSeek-R1-052892.3%

SimpleQA Score Distribution

A closer view of the leading scores on this benchmark.

SimpleQA

The Top AI Models for SimpleQA

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

Ranking basisThis simpleqa 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
    DE
    DeepSeek-V3.2-ExpDeepSeek
    Score
    97.1%
    Price
    $0.14 input / $0.28 output per 1M tokens
    Speed
    Up to 97 tok/s via DeepInfra

    Strengths

    • Ranks #1 of 47 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SimpleQA, not total model capability
  2. 02
    XA
    Grok 4 FastxAI
    Score
    95.0%
    Speed
    Up to 90 tok/s via xAI

    Strengths

    • Ranks #2 of 47 compared models
    • 98th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SimpleQA, not total model capability
  3. 03
    DE
    DeepSeek-V3.1DeepSeek
    Score
    93.4%

    Strengths

    • Ranks #3 of 47 compared models
    • 96th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SimpleQA, not total model capability
  4. 04
    DE
    DeepSeek-R1-0528DeepSeek
    Score
    92.3%
    Speed
    Up to 45 tok/s via DeepInfra

    Strengths

    • Ranks #4 of 47 compared models
    • 93th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SimpleQA, not total model capability
  5. 05
    BA
    ERNIE 5.0Baidu
    Score
    75.0%

    Strengths

    • Ranks #5 of 47 compared models
    • 91th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures SimpleQA, not total model capability

Selection summary

Best AI Models for SimpleQA

DeepSeek-V3.2-Exp currently leads SimpleQA with 97.1%. 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 #1DeepSeek-V3.2-Exp97.1% · $0.14 input / $0.28 output per 1M tokensBenchmark rank #2Grok 4 Fast95.0% · Up to 90 tok/s via xAIBenchmark rank #3DeepSeek-V3.193.4%

What is SimpleQA?

What SimpleQA measures and how its scores work.

SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.

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

Family
SimpleQA
Modality
text
Primary category
reasoning
Score direction
higher
LLMBoard eligible
Yes
Evaluation key
simpleqa|llm-stats-current

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

FAQ

Common questions about SimpleQA.

Which model scores highest on SimpleQA?

DeepSeek-V3.2-Exp is currently ranked first with 97.1%.

What does SimpleQA measure?

SimpleQA is a factuality benchmark developed by OpenAI that measures the short-form factual accuracy of large language models. The benchmark contains 4,326 short, fact-seeking questions that are adversarially collected and designed to have single, indisputable answers. Questions cover diverse topics from science and technology to entertainment, and the benchmark also measures model calibration by evaluating whether models know what they know.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

47 model results are currently shown.

Does this benchmark affect the overall score?

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