Kimi K3
Kimi open 2026-07-16
SOTA
Kimi K3 outscores the executioner, Claude Haiku 5.5 (43.4): it is SOTA, the top tier right now.
- Score
- 43.6
- Artificial Analysis Intelligence Index
- Best rank
- #14
- Arena — Code
- Price
- $6.00
- Blended Price, USD per 1M tokens (3:1 input:output)
- Best effort
- max
- Artificial Analysis LLM Leaderboard
Every score this model has on the boards we track, one board at a time. Nothing is averaged across boards.
Scores on each board Each rank is the board’s own: the rank it publishes, or the model’s place in the board’s own order where it publishes none. The bar shows how much of that board the model ranks ahead of. Scores from different boards cannot be compared.
- Artificial Analysis LLM LeaderboardArtificial Analysis Intelligence Index43.6No. 18 of 375
- Arena — AgentOverall0.038No. 17 of 52
- Artificial Analysis Coding AgentsArtificial Analysis Coding Agent Index0.519No. 15 of 24
- Epoch AI Benchmarking HubFrontierMath Tiers 1–372.2%No. 22 of 82
- HarnessTax — SWE-bench LiteSWE-bench Lite76.7%No. 4 of 7
- HarnessTax — Terminal-Bench 2.0Terminal-Bench 2.073.3%No. 4 of 7
- Arena — CodeOverall1654No. 14 of 142
- Arena — TextOverall (style control)1488No. 16 of 414
Scores at each effort setting The same model at each effort setting. A bar runs from zero to that board’s best score, so read each board on its own and never across boards.
Artificial Analysis LLM Leaderboardbest at max
- low30.1
- max43.6
Tested at one setting only
- Arena — Agent · agentmax0.038
- Epoch AI Benchmarking Hubmax72.2%
- Arena — Codemax1654
- Arena — Textmax1488
Value: where it sits on the kill line Each dot is a model: further right costs more, higher up scores more. Where the dashed lines cross stands the executioner. Above it is SOTA, the top tier; below and to the left is cheaper and weaker, the low-cost picks.
The executioner is the best value on the board: every model that costs more and scores no higher is killed.
Compare on the home page chart → How the kill line is drawn →
Providers
Where to run it: the developer's own page first, then the clouds that serve it.
- Kimi Official kimi.ai ↗
- Amazon Bedrock docs.aws.amazon.com ↗
- Microsoft Foundry ai.azure.com ↗
- Alibaba Cloud Model Studio alibabacloud.com ↗
- Tencent Cloud intl.cloud.tencent.com ↗
- OpenRouter openrouter.ai ↗
Other Kimi models
Price, and the Artificial Analysis Intelligence Index score from Artificial Analysis LLM Leaderboard.
- Kimi K2.6 Killed $1.71 27.0
- Kimi K2.7 Code Killed $1.71 25.8
- Kimi K2.5 Killed $1.14 23.5
- Kimi Linear Instruct — 7.3
Compare with another SOTA model
- Kimi K3 vs Qwen 3.8 Max
- Kimi K3 vs Claude Fable 5
- Kimi K3 vs Claude Fable 5.1
- Kimi K3 vs Claude Haiku 5.5
- Kimi K3 vs Claude Opus 5.5
- Kimi K3 vs Claude Sonnet 5.5
- Kimi K3 vs Gemini 4 Argon
- Kimi K3 vs Muse Spark 1.3
- Kimi K3 vs GPT 6.1 Sol
- Kimi K3 vs GPT 6 Astra
- Kimi K3 vs Step 5
- Kimi K3 vs Grok 4.7
- Kimi K3 vs MiMo V2.6 Pro
- Kimi K3 vs GLM 5.3
Details
- Family
- Kimi
- Blended Price, USD per 1M tokens (3:1 input:output)
- $6.00
- Sources
- 8 boards ·13 scores ·7 settings
- Also written as
-
Kimi K3kimi-k3moonshotai/Kimi-K3