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
1367 models808 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

reasoning benchmark

LiveCodeBench Leaderboard

LiveCodeBench is a benchmark for evaluating large language models on coding problems collected from LeetCode, AtCoder, and CodeForces.

Updated Oct 23, 2026

Models76
Model coverage76
MetricScore
EvidenceC

On this page

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

LiveCodeBench Ranking

Higher score ranks better on this benchmark.

30 of 76 rows
Columns

Show columns

Sort by
Rank
Model
Score
Percentile
Participants
Evidence
Evaluated
Rank01ModelDEDeepSeek-V4-Pro-MaxDeepSeekScore93.50%Percentile100.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank02ModelDEDeepSeek-V4-Flash-MaxDeepSeekScore91.60%Percentile98.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank03ModelDEDeepSeek-V4-Flash-0423DeepSeekScore88.40%Percentile97.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank04ModelRABeam 1.0Reflection AIScore88.30%Percentile96.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank05ModelUPSolar Pro 4UpstageScore87.80%Percentile94.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank06ModelDEDeepSeek-V3.2 (Thinking)DeepSeekScore83.30%Percentile93.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank07ModelDEDeepSeek-V3.2DeepSeekScore83.30%Percentile92.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank08ModelMIMiniMax M2MiniMaxScore83.00%Percentile90.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank09ModelMELongCat-Flash-Thinking-2601MeituanScore82.80%Percentile89.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank10ModelNVNemotron 3 Super (120B A12B)NVIDIAScore81.19%Percentile88.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank11ModelXAGrok-3 MinixAIScore80.40%Percentile86.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank12ModelXAGrok 4 FastxAIScore80.00%Percentile85.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank13ModelXAGrok-3xAIScore79.40%Percentile84.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank14ModelXAGrok-4 HeavyxAIScore79.40%Percentile82.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank15ModelMELongCat-Flash-ThinkingMeituanScore79.40%Percentile81.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank16ModelXAGrok-4xAIScore79.00%Percentile80.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank17ModelMIMiniMax M2.1MiniMaxScore78.00%Percentile78.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank18ModelAMNova 2 ProAmazonScore74.60%Percentile77.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank19ModelDEDeepSeek-V3.2-ExpDeepSeekScore74.10%Percentile76.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank20ModelDEDeepSeek-R1-0528DeepSeekScore73.30%Percentile74.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank21ModelZAGLM-4.5Zhipu AIScore72.90%Percentile73.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank22ModelNVNemotron Nano 9B v2NVIDIAScore71.10%Percentile72.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank23ModelAMNova 2 LiteAmazonScore71.00%Percentile70.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank24ModelZAGLM-4.5-AirZhipu AIScore70.70%Percentile69.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank25ModelACQwen3 235B A22BAlibaba Cloud / Qwen TeamScore70.70%Percentile68.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank26ModelGOGemini 2.5 Pro Preview 06-05GoogleScore69.00%Percentile66.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank27ModelINMercury 2InceptionScore67.00%Percentile65.33%Participants76EvidenceCEvaluatedOct 22, 2026
Rank28ModelNVLlama 3.1 Nemotron Ultra 253B v1NVIDIAScore66.31%Percentile64.00%Participants76EvidenceCEvaluatedOct 22, 2026
Rank29ModelACQwen3 32BAlibaba Cloud / Qwen TeamScore65.70%Percentile62.67%Participants76EvidenceCEvaluatedOct 22, 2026
Rank30ModelMIMiniMax M1 80KMiniMaxScore65.00%Percentile61.33%Participants76EvidenceCEvaluatedOct 22, 2026

LiveCodeBench Highlights

The leading models and scores on this benchmark.

LiveCodeBench Score Distribution

A closer view of the leading scores on this benchmark.

LiveCodeBench

The Top AI Models for LiveCodeBench

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

What is LiveCodeBench?

What LiveCodeBench measures and how its scores work.

LiveCodeBench continuously collects programming contest problems and annotates them with release dates to support evaluation on problems released after a model's training cutoff.

It measures Score, reported as a ratio, across code generation, self-repair, code execution, and test output prediction.

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

Family
LiveCodeBench
Modality
text
Primary category
reasoning
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 LiveCodeBench.

Which model scores highest on LiveCodeBench?

DeepSeek-V4-Pro-Max is currently ranked first with 93.50%.

What are the top three models on LiveCodeBench?

The current leaders are DeepSeek-V4-Pro-Max (93.50%), DeepSeek-V4-Flash-Max (91.60%), and DeepSeek-V4-Flash-0423 (88.40%).

Which LiveCodeBench model has the lowest official input price?

Min istral 3 (3B Reasoning 2512) has the lowest matched official input price at $0.10 input / $0.10 output per 1M tokens.

Which models are fastest among LiveCodeBench results?

The fastest matched records are Min istral 3 (3B Reasoning 2512) (1,222.00 tok/s via Mistral AI), Mercury 2 (1,009.00 tok/s via Inception), and Llama 4 Scout (776.10 tok/s via Groq).

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 LiveCodeBench measure?

It measures Score, reported as a ratio, across code generation, self-repair, code execution, and test output prediction.

Is a higher score better?

Yes. Higher values rank better for this benchmark.

How many models are compared?

76 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 #1DeepSeek-V4-Pro-Max93.50%
Rank #2DeepSeek-V4-Flash-Max91.60%
Rank #3DeepSeek-V4-Flash-042388.40%
Rank #4Beam 1.088.30%

Ranking basisThis livecodebench 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
    DE
    DeepSeek-V4-Pro-MaxDeepSeek
    Score
    93.50%
    Price
    $0.66 input / $2.0 output per 1M tokens

    Strengths

    • Ranks #1 of 76 compared models
    • 100th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures LiveCodeBench, not total model capability
  2. 02
    DE
    DeepSeek-V4-Flash-MaxDeepSeek
    Score
    91.60%
    Price
    $0.15 input / $0.60 output per 1M tokens

    Strengths

    • Ranks #2 of 76 compared models
    • 99th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures LiveCodeBench, not total model capability
  3. 03
    DE
    DeepSeek
    Score
    88.40%
    Price
    $0.15 input / $0.60 output per 1M tokens

    Strengths

    • Ranks #3 of 76 compared models
    • 97th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures LiveCodeBench, not total model capability
  4. 04
    RA
    Reflection AI
    Score
    88.30%

    Strengths

    • Ranks #4 of 76 compared models
    • 96th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures LiveCodeBench, not total model capability
  5. 05
    UP
    Upstage
    Score
    87.80%
    Price
    $0.30 input / $1.2 output per 1M tokens

    Strengths

    • Ranks #5 of 76 compared models
    • 95th percentile on this benchmark
    • C evidence result

    Considerations

    • This result measures LiveCodeBench, not total model capability

Selection summary

Best AI Models for LiveCodeBench

DeepSeek-V4-Pro-Max currently leads LiveCodeBench with 93.50%. 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.

DeepSeek-V4-Flash-0423
Beam 1.0
Solar Pro 4
Benchmark rank #1DeepSeek-V4-Pro-Max93.50% · $0.66 input / $2.0 output per 1M tokens
Benchmark rank #2DeepSeek-V4-Flash-Max91.60% · $0.15 input / $0.60 output per 1M tokens
Benchmark rank #3DeepSeek-V4-Flash-042388.40% · $0.15 input / $0.60 output per 1M tokens