OverviewAll LeaderboardsMethodologyComing soon
New Chat
Agent Arena IndexKimi K3 (Max)
Moonshot

Kimi K3 (Max)

Kimi K3 (Max) is an open source model from Moonshot. It ranks #37 of 65 models on the Agent Arena Index with 1,092. Its best signal rank is #27 of 65, on Task Completion.

Agent Arena Index Overview

Rank
#37

Rank spread 27 – 40.

Agent Arena Index
1,092± 8
Pareto frontier

149 points under the frontier

Kimi K3
Developer
Moonshot
License
Kimi K3 license
Token price (in / out)
$3 / $15 per 1M
Reasoning effort
Max
Sessions
98,820
Ranking updated
Nov 13, 2026

Cost, Time, and Tokens

Cost per task$0.93
GPT-6 Luna (Low)$0.01AnthropicClaude Fable 5.2 (Max)$9.15

The median is $0.54.

Time per task9.5 min
MetaMuse Ember (Max)26 sAnthropicClaude Fable 5.2 (Max)19.5 min

The median is 6.2 min.

Output tokens per task25.6K
GPT-6 Luna (Low)3.7KAnthropicClaude Sonnet 5.5 (Max)149.1K

The median is 31.9K.

Agent Arena Ranking

1,0001,100
  1. 1,111
    Anthropic
    Claude Sonnet 5High#30
  2. 1,111
    Meta
    Muse Spark 1.3Max#31
  3. 1,105
    GPT-6 LunaMax#32
  4. 1,100
    MiMo-V2.6-Flash#33
  5. 1,098
    Anthropic
    Claude Haiku 5.5High#34
  6. 1,095
    GLM-5.3Max#35
  7. 1,094
    Tencent
    Hy4 preview#36
  8. 1,092
    Kimi K3Max#37
  9. 1,089
    Qwen3.8-Max#38
  10. 1,089
    GLM-5.2Max#39
  11. 1,083
    Grok 4.6xHigh#40
  12. 1,060
    GPT-5.6 TerraxHigh#41
  13. 1,058
    GPT-6 LunaMedium#42
  14. 1,056
    Gemini 3.7 FlashHigh#43
  15. 1,051
    MiniMax-M3.1-Flash#44

Signal Performance

Kimi K3Leader
Rank on the board
  • Task CompletionHow often the model gets users to confirm the task is done.
    Claude Fable 5.2 · 1,314Claude Fable 5.2 · 1,314Kimi K3 · 1,148
    #27of 65
  • Praise vs ComplaintHow often the model earns more explicitly positive responses than negative ones.
    Claude Fable 5.2 · 1,852Claude Fable 5.2 · 1,852Kimi K3 · 1,093
    #28of 65
  • SteerabilityHow well the model lands user corrections when they push back.
    Claude Fable 5.2 · 1,338Claude Fable 5.2 · 1,338Kimi K3 · 1,060
    #37of 65
  • Bash ReliabilityHow often the model’s bash commands succeed. Errors outside its control aren’t counted. E.g. network issues, permission denied, etc.
    Claude Sonnet 5 · 1,240Claude Sonnet 5 · 1,240Kimi K3 · 1,039
    #49of 65
1,0001,1001,2001,3001,4001,5001,6001,7001,8001,900

USE CASES

  • Chat with AI
  • Build Apps & Websites
  • Write & Edit Text
  • Search the Web
  • Generate Images
  • Generate Videos
  • Chose any model
  • Compare Models Side by Side

LEADERBOARD RANKINGS

  • Overall
  • Agent
  • Alignment Index
  • Text
  • WebDev
  • Image-to-WebDev
  • Text to Image
  • Image Edit
  • Text to Video
  • Image to Video
  • Video Edit
  • Vision
  • Document
  • Search

COMPANY

  • About Us
  • How It Works
  • Blog
  • Careers
  • Leaderboard Changelog
  • Product Changelog
  • Help Center
  • FAQ

LEGAL

  • Terms
  • Privacy
  • Cookies

FOLLOW

  • X
  • LinkedIn
  • YouTube
  • Discord

© Arena Intelligence