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

    Qwen: Qwen3 Next 80B A3B Thinking

    Released Sep 11, 2025
    qwen/qwen3-next-80b-a3b-thinking
    text→text
    Set alert
    Ranked in:Text
    Try it on:
    OOpenRouter

    Qwen3-Next-80B-A3B-Thinking is a reasoning-first chat model in the Qwen3-Next line that outputs structured “thinking” traces by default. It’s designed for hard multi-step problems; math proofs, code synthesis/debugging, logic, and agentic...

    Best overall rank
    #147
    Across all subsets · overall
    Total arena votes
    74,954
    Across 27 (subset, category) entries
    Providers on OR
    5
    Nebius + others
    First seen
    2026-05-08
    26 days tracked

    Trend

    Pricing

    Per token

    Input (per 1M)
    $0.098
    Output (per 1M)
    $0.780

    Capabilities

    Modality
    text->text
    Context
    262K tokens
    Tokenizer
    Qwen3
    Knowledge cutoff
    2025-09-30
    Max output
    33K tokens
    Moderation
    Unmoderated
    Hugging Face
    Qwen/Qwen3-Next-80B-A3B-Thinking
    Supported parameters
    frequency_penalty
    include_reasoning
    logit_bias
    max_tokens
    min_p
    presence_penalty
    reasoning
    repetition_penalty
    response_format
    seed
    stop
    structured_outputs
    temperature
    tool_choice
    tools
    top_k
    top_p

    Providers

    ProviderRegion / TagContextInput/MOutput/MUptime 1dLatencyQuantCache
    Nebiusnebius/fp8128K$0.150$1.20100.00%-
    fp8
    -
    Novitanovita/bf16131K$0.150$1.50100.00%-
    bf16
    -
    Alibabaalibaba131K$0.098$0.78099.96%---
    AtlasCloudatlas-cloud/fp8262K$0.150$1.5099.03%-
    fp8
    -
    Googlegoogle-vertex/global262K$0.150$1.2087.09%---

    Per-category ranks

    Text

    RankCategoryVariantTrendEloVotes
    #122Japanese
    thinking
    1296180
    #123German
    thinking
    1366287
    #125Chinese
    thinking
    1413718
    #128Korean
    thinking
    1319332
    #128Math
    thinking
    1389829
    #132Industry Mathematical
    thinking
    1387671
    #141Coding
    thinking
    14212,677
    #141English
    thinking
    13876,634
    #141Expert
    thinking
    1385620
    #142Industry Software And It Services
    thinking
    14104,835
    #143Industry Medicine And Healthcare
    thinking
    1393721
    #144French
    thinking
    1369195
    #147Exclude Ties
    thinking
    13389,762
    #147Instruction Following
    thinking
    13593,514
    #147Overall
    thinking
    137013,700
    #148Hard Prompts
    thinking
    13856,693
    #148Industry Business And Management And Financial Operations
    thinking
    13672,497
    #149Spanish
    thinking
    1350398
    #150Longer Query
    thinking
    13702,831
    #151Hard Prompts English
    thinking
    13973,495
    #151Russian
    thinking
    1353750
    #152Industry Life And Physical And Social Science
    thinking
    13802,125
    #158Multi Turn
    thinking
    13502,296
    #159Creative Writing
    thinking
    13241,796
    #159Industry Writing And Literature And Language
    thinking
    13393,077
    #163Industry Entertainment And Sports And Media
    thinking
    13172,453
    #169Industry Legal And Government
    thinking
    1360868