AnotherWrapper
    $249
    LLMImageVideoSpeechTranscription

    LiveCodeBench

    🏆 Leaderboard

    As of August 20, 2026, DeepSeek-V4-Pro-Max is #1 for LiveCodeBench at 93.5%. Ranked by LiveCodeBench: contest programming on problems released after training. LiveCodeBench leaderboard for contamination-resistant coding scores, plus API pricing. Rank LLMs on fresh programming problems.

    Updated August 20, 2026282 models33 providers
    DeepSeek-V4-Pro-MaxOSS
    DeepSeek · Open Source
    		93.5%———$5.22
    DeepSeek-V4-Flash-MaxOSS
    DeepSeek · Open Source
    		91.6%———$0.42
    Claude Fable 5
    Anthropic · Proprietary
    		89.8%———$60.00
    Claude Opus 5
    Anthropic · Proprietary
    		89.0%———$30.00
    Gemini 3.7 Flash
    Google · Proprietary
    		88.7%———$4.50
    Gemini 3.1 Pro
    Google · Proprietary
    		88.5%———$17.50
    DeepSeek-V4-Flash-0423OSS
    DeepSeek · Open Source
    		88.4%———$0.30
    Grok 4.6
    xAI · Proprietary
    		88.2%———$8.00
    Gemini 3.6 Flash
    Google · Proprietary
    		88.1%———$4.50
    GPT-5.2 Codex
    OpenAI · Proprietary
    		88.0%———$15.75
    Qwen3.8 MaxOSS
    Qwen · Open Source
    		87.8%———$8.00
    Claude Opus 4.8
    Anthropic · Proprietary
    		87.8%———$30.00
    Solar Pro 4
    Upstage · Proprietary
    		87.8%———$1.50
    Gemini 3.5 Flash
    Google · Proprietary
    		87.6%———$10.50
    DeepSeek-V4-Pro-0813OSS
    DeepSeek · Open Source
    		87.5%———$5.28
    Grok 4.5
    xAI · Proprietary
    		87.3%———$8.00
    GPT-5.3 Codex
    OpenAI · Proprietary
    		87.3%———$15.75
    DeepSeek-V4-Flash-0731OSS
    DeepSeek · Open Source
    		87.3%———$1.76
    Kimi K3OSS
    Moonshot AI · Open Source
    		87.2%———$18.00
    Qwen3.7 Max
    Qwen · Proprietary
    		87.1%—91.6%—$5.00
    Kimi K2.6OSS
    Moonshot AI · Open Source
    		86.8%—89.6%—$4.93
    GPT-5 mini
    OpenAI · Proprietary
    		86.6%———$2.25
    GPT-5.1
    OpenAI · Proprietary
    		86.5%———$11.25
    Gemini 3 Pro
    Google · Proprietary
    		86.4%———$14.00
    Nemotron 3 Ultra (550B A55B)OSS
    NVIDIA · Open Source · via OpenRouter
    		86.0%—89%—$2.70
    Qwen3.6 Plus
    Qwen · Proprietary
    		86.0%—87.1%—$3.50
    Inkling-SmallOSS
    Thinking Machines · Open Source · via Thinking Machines Lab
    		85.9%———$1.50
    GPT-5.6 Terra
    OpenAI · Proprietary
    		85.9%———$14.00
    Muse Spark 1.1
    Meta · Proprietary
    		85.9%———$5.50
    GPT-5
    OpenAI · Proprietary
    		85.9%93.4%—88%$11.25
    Gemini 3 Flash
    Google · Proprietary
    		85.6%———$3.50
    GPT-5.1 Codex
    OpenAI · Proprietary
    		85.5%———$11.25
    GPT-5.2
    OpenAI · Proprietary
    		85.4%———$15.75
    GPT-5.5
    OpenAI · Proprietary
    		85.3%———$35.00
    Claude Opus 4.7
    Anthropic · Proprietary
    		85.1%———$30.00
    GPT-5 Codex
    OpenAI · Proprietary
    		84.7%———$11.25
    Grok 4.3
    xAI · Proprietary
    		84.5%———$3.75
    GPT-5.4
    OpenAI · Proprietary
    		84.1%———$17.50
    GPT-5.4 nano
    OpenAI · Proprietary
    		84.0%———$1.45
    o3
    OpenAI · Proprietary
    		83.9%——81.3%$10.00
    DeepSeek-V3.2OSS
    DeepSeek · Open Source · via OpenRouter
    		83.3%———$0.57
    GPT OSS 120BOSS
    OpenAI · Open Source
    		83.2%———$0.54
    MiniMax M2OSS
    MiniMax · Open Source
    		83%———$1.50
    LongCat-Flash-Thinking-2601OSS
    Meituan · Open Source
    		82.8%———$1.50
    GPT-5.6 Sol
    OpenAI · Proprietary
    		82.6%———$35.00
    Claude Sonnet 5
    Anthropic · Proprietary
    		82.4%———$12.00
    GLM-4.7OSS
    Z AI · Open Source
    		82.2%84.9%84.9%—$2.80
    o4-mini
    OpenAI · Proprietary
    		82.2%——68.9%$5.50
    MiniMax M3OSS
    MiniMax · Open Source
    		82.2%———$1.50
    Claude Sonnet 4.6
    Anthropic · Proprietary
    		82.1%———$18.00
    Showing 1–50 of 282 models

    Next step

    You found the model. Now ship the product.

    Auth, billing, and the API layer are already decided. Fork 8 finished AI apps, or have us build the first version with you.

    Get the stack — $249Have us build it with you

    The index

    All Large Language Models

    282 models across 33 providers. Search or jump to a lab — every model page stays linked here.

    OpenAI51Anthropic22Google25Cursor2xAI15DeepSeek16MiniMax6Qwen38AI21 Labs2Amazon5Baidu1ByteDance2Cohere3IBM1Inception1inclusionAI1LG AI Research1Meituan4Meta12Microsoft5Mistral23Moonshot AI9Nous Research1NVIDIA5Poolside2Sakana AI1StepFun1Tencent1Thinking Machines1Unisound1Upstage1Xiaomi3Z AI20

    Baidu

    1 models
    ERNIE 4.5

    Inception

    1 models
    Mercury 2

    inclusionAI

    1 models
    Ling-3.0-flash

    LG AI Research

    1 models
    K-EXAONE-236B-A23B

    Nous Research

    1 models
    Hermes 3 70B

    Sakana AI

    1 models
    Sakana Namazu

    StepFun

    1 models
    Step-3.5-Flash

    Tencent

    1 models
    Hy3

    Thinking Machines

    1 models
    Inkling-Small

    Unisound

    1 models
    U2

    Upstage

    1 models
    Solar Pro 4

    Which model leads LiveCodeBench right now?

    As of August 20, 2026, DeepSeek-V4-Pro-Max by DeepSeek is #1 for LiveCodeBench at 93.5%. Ranked by LiveCodeBench: contest programming on problems released after training. This board also tracks LiveCodeBench, HumanEval, LiveCodeBench v6, Alder Polyglot. Next on the same board: DeepSeek-V4-Flash-Max and Claude Fable 5. Related leaders: Kimi K2 0905 on HumanEval at 94.5%; Qwen3.7 Max on LiveCodeBench v6 at 91.6%. This livecodebench leaderboard ranks models by LiveCodeBench. Scores come from public evals. Prices are the live API rates in the table above.

    Sources: LiveCodeBench (https://livecodebench.github.io/); OpenAI API pricing (https://developers.openai.com/api/docs/pricing); Anthropic Claude API pricing (https://platform.claude.com/docs/en/about-claude/pricing); Gemini API pricing (https://ai.google.dev/gemini-api/docs/pricing)

    Top 8 for livecodebench. Ranked by LiveCodeBench: contest programming on problems released after training. Input and output are dollars per million tokens.
    RankModelLiveCodeBenchInput /MOutput /M
    1DeepSeek-V4-Pro-Max93.5%$1.74$3.48
    2DeepSeek-V4-Flash-Max91.6%$0.14$0.28
    3Claude Fable 589.8%$10.00$50.00
    4Claude Opus 589.0%$5.00$25.00
    5Gemini 3.7 Flash88.7%$0.75$3.75
    6Gemini 3.1 Pro88.5%$2.50$15.00
    7DeepSeek-V4-Flash-042388.4%$0.10$0.20
    8Grok 4.688.2%$2.00$6.00

    Related LLM leaderboards

    LLM LeaderboardCodingMathAgenticReasoningWritingScienceFinanceLong ContextVisionOpen Source

    Individual benchmark pages

    Rank one eval at a time. All LLM benchmarks.

    HLESWE BenchHumanity's Last Exam (No Tools)Humanity's Last Exam (With Tools)Terminal-Bench 2.0Terminal-BenchTerminal-Bench 2.1Terminal-Bench 2Terminal-Bench 3.0ARC-AGI-1 Verified

    LiveCodeBench FAQ

    Who ranks #1 on the LiveCodeBench leaderboard?

    As of August 20, 2026, DeepSeek-V4-Pro-Max by DeepSeek ranks #1 on LiveCodeBench at 93.5%. API pricing is $1.74/M input and $3.48/M output.

    What are the top models on LiveCodeBench?

    The current LiveCodeBench ranking as of August 20, 2026 is 1. DeepSeek-V4-Pro-Max at 93.5%; 2. DeepSeek-V4-Flash-Max at 91.6%; 3. Claude Fable 5 at 89.8%.

    Which livecodebench model is the cheapest?

    Gemma 3 4B is the cheapest scored model on this livecodebench leaderboard at $0.02/M input and $0.04/M output ($0.06 blended). DeepSeek-V4-Pro-Max still leads LiveCodeBench at 93.5%.

    Should I always pick the #1 LiveCodeBench model?

    Not automatically. DeepSeek-V4-Pro-Max leads LiveCodeBench, but a cheaper scored model can be the better production choice if the quality gap is small. Use the table to weigh LiveCodeBench against input/output price, context window, and related evals.

    How often is the LiveCodeBench leaderboard updated?

    Scores and API prices on this page are refreshed from published evals and provider rates. The snapshot is labeled August 20, 2026. Treat it as a current index, not a one-off blog post.

    What is LiveCodeBench?

    LiveCodeBench scores models on new competitive-programming problems released after training cutoffs, which reduces memorization compared with HumanEval.

    How is LiveCodeBench different from SWE-bench?

    LiveCodeBench is contest-style code. SWE-bench is multi-file GitHub repair. Use LiveCodeBench for algorithms, SWE-bench for product engineering.

    Why show HumanEval next to LiveCodeBench?

    HumanEval is older and more saturated, but still widely cited. Seeing both columns makes it obvious when a model only looks strong on the easy set.