gpt·basedElo-ranked, value-sorted
Sign in

Stay updated

Get an email when a cheaper, stronger model than yours appears on the leaderboard. No spam.

gpt·based

The LLM leaderboard that tells you when to switch. Arena rankings joined with OpenRouter pricing, refreshed daily.

Use cases

  • Best for coding
  • Best for multilingual
  • Cheapest worth using
  • Compare two models

Browse

  • Text models
  • WebDev models
  • Vision models
  • Image generation
  • Image edit

Data sources

  • LMArena
  • Providers

Developers

  • API on RapidAPI
  • OpenAPI spec

Contact

  • Contact page
  • [email protected]
© 2026 gptbasedDaily refresh, never paid placement.
    gpt·basedElo-ranked, value-sorted
    Sign in
    Back to WebDev · Overall

    DeepSeek: DeepSeek V3.2

    Released Dec 1, 2025
    deepseek/deepseek-v3.2
    text→text
    Set alert
    Ranked in:TextWebDev
    Try it on:
    OOpenRouter

    DeepSeek-V3.2 is a large language model designed to harmonize high computational efficiency with strong reasoning and agentic tool-use performance. It introduces DeepSeek Sparse Attention (DSA), a fine-grained sparse attention mechanism...

    Best overall rank
    #54
    ▼ 1 vs 7d ago
    Total arena votes
    373,928
    Across 62 (subset, category) entries
    Providers on OR
    11
    Baidu + others
    First seen
    2026-05-08
    45 days tracked

    Trend

    Pricing

    Per token

    Input (per 1M)
    $0.229
    Output (per 1M)
    $0.343

    Capabilities

    Modality
    text->text
    Context
    131K tokens
    Tokenizer
    DeepSeek
    Max output
    64K tokens
    Moderation
    Unmoderated
    Hugging Face
    deepseek-ai/DeepSeek-V3.2
    Supported parameters
    frequency_penalty
    include_reasoning
    logit_bias
    logprobs
    max_tokens
    min_p
    presence_penalty
    reasoning
    repetition_penalty
    response_format
    seed
    stop
    structured_outputs
    temperature
    tool_choice
    tools
    top_k
    top_logprobs
    top_p

    Providers

    ProviderRegion / TagContextInput/MOutput/MUptime 1dLatencyQuantCache
    Baidubaidu/fp8131K$0.252$0.37899.98%-
    fp8
    -
    Friendlifriendli164K$0.500$1.5099.92%---
    StreamLakestreamlake128K$0.229$0.34399.37%---
    Novitanovita/fp8164K$0.269$0.40099.36%-
    fp8
    -
    Alibabaalibaba131K$0.370$1.1199.27%---
    AtlasCloudatlas-cloud/fp8164K$0.260$0.38099.20%-
    fp8
    -
    SiliconFlowsiliconflow/fp8164K$0.259$0.42099.18%-
    fp8
    -
    DigitalOceandigitalocean66K$0.425$1.3699.15%---
    DeepInfradeepinfra/fp4164K$0.260$0.38098.63%-
    fp4
    -
    SambaNovasambanova33K$3.00$4.5098.27%---
    Googlegoogle-vertex164K$0.560$1.6897.89%---

    Per-category ranks

    Text

    RankCategoryVariantTrendEloVotes
    #59Japanese
    thinking
    1390311
    #63Hard Prompts English
    thinking
    146210,612
    #69Industry Writing And Literature And Language
    -
    141010,401
    #70Industry Mathematical
    thinking
    1440393
    #71Longer Query
    -
    144212,953
    #72English
    thinking
    14414,365
    #73Creative Writing
    -
    14016,687
    #74Coding
    thinking
    14751,919
    #77Expert
    thinking
    1448396
    #77Hard Prompts
    -
    144726,212
    #77Spanish
    -
    14171,274
    #78Chinese
    thinking
    1465304
    #78Industry Entertainment And Sports And Media
    -
    13968,542
    #78Instruction Following
    -
    14173,310
    #78Math
    thinking
    14262,506
    #79Chinese
    -
    14642,442
    #79Instruction Following
    thinking
    14172,544
    #79Korean
    -
    1368762
    #80Hard Prompts
    thinking
    144622,470
    #80Industry Business And Management And Financial Operations
    -
    14218,818
    #81English
    -
    143721,521
    #81Exclude Ties
    -
    141633,503
    #81Hard Prompts English
    -
    14543,206
    #82French
    -
    14421,118
    #82Industry Software And It Services
    thinking
    14593,227
    #82Multi Turn
    -
    14288,265
    #83German
    thinking
    1411719
    #84German
    -
    1411861
    #85Overall
    -
    142311,922
    #86Overall
    thinking
    142341,085
    #87Coding
    -
    14662,499
    #87Exclude Ties
    thinking
    141228,813
    #87Industry Life And Physical And Social Science
    thinking
    14386,583
    #87Longer Query
    thinking
    14312,074
    #88Creative Writing
    thinking
    13926,206
    #88Math
    -
    1418775
    #88Multi Turn
    thinking
    14231,436
    #91Industry Entertainment And Sports And Media
    thinking
    13897,504
    #91Industry Software And It Services
    -
    14544,132
    #93Industry Writing And Literature And Language
    thinking
    13972,096
    #94Industry Life And Physical And Social Science
    -
    14351,910
    #95Japanese
    -
    1347120
    #96Industry Legal And Government
    thinking
    1423521
    #96Industry Medicine And Healthcare
    -
    1436583
    #99Russian
    -
    1405844
    #104Industry Medicine And Healthcare
    thinking
    14332,709
    #104Russian
    thinking
    14024,508
    #107Expert
    -
    1422609
    #107French
    thinking
    1421139
    #107Industry Business And Management And Financial Operations
    thinking
    14051,799
    #107Industry Legal And Government
    -
    1416782
    #109Industry Mathematical
    -
    1413633
    #109Korean
    thinking
    1345195
    #114Spanish
    thinking
    1386277

    WebDev

    RankCategoryVariantTrendEloVotes
    #49Webdev-react
    thinking
    13623,896
    #52Webdev-html
    thinking
    13723,998
    #53Webdev
    thinking
    13687,910
    #54Overall
    thinking
    13687,911
    #56Webdev-react
    -
    13475,199
    #68Webdev
    -
    12874,873
    #72Overall
    -
    12874,873
    #73Webdev-html
    -
    12954,873

    Related models

    DeepSeek: DeepSeek V3.2 Expdeepseek/deepseek-v3.2-exp