Skip to main content

Kimi K2

KimiKimiOpen WeightMIT · Commercial OK

Description

Kimi K2-Instruct-0905 is the latest, most capable version of Kimi K2, achieving state-of-the-art performance in frontier knowledge, math, and coding among non-thinking models. This Mixture-of-Experts model features 32 billion activated parameters and 1 trillion total parameters, meticulously optimized for agentic tasks. Key features include enhanced agentic coding intelligence, extended context length to 256K tokens, and a hybrid architecture trained with MuonClip optimizer on 15.5T tokens. The model achieves 65.8% on SWE-bench Verified (single attempt), 47.3% on SWE-bench Multilingual, and excels at tool use with 70.6% on Tau2-retail. It is a reflex-grade model without long thinking, designed to act and execute complex tasks seamlessly.

Release Date
2025-07-11
Parameters
1.0T
Context Length
262K
Modalities
text

Capability Radar

60
general
50
coding
70
reasoning
68
scienceest.
60
agents
0
multimodal

Science is estimated from LLM Stats science scores or reasoning when no dedicated science benchmarks are available.

Rankings

No ranking data available

Benchmark Scores (LLM Stats)

(LLM Stats (zeroeval))

Chat

IFEvalGoogle Research (2023)89.8%SR
Tau2 Airline56.5%SR
Multi-Challenge54.1%SR

Communication

Tau2 Retail70.6%SR
Tau2 Telecom65.8%SR

Factuality

SimpleQA31.0%SR

General

C-Eval92.5%SR
MMLU89.5%SR
MultiPL-E85.7%SR
TriviaQA85.1%SR
CSimpleQA78.4%SR
ACEBench76.5%SR
Aider-Polyglot60.0%SR
SWE-bench Verified (Agentless)51.8%SR

Language

MMLU-Redux92.7%SR
MMLU-redux-2.090.2%SR
MMLU-Pro81.1%SR

Math

MATH-50097.4%SR
GSM8k97.3%SR
CBNSL95.6%SR
LiveBench76.4%SR
CNMO 202474.3%SR
MATH70.2%SR
AIME 202469.6%SR
PolyMath-en65.1%SR
AIME 202549.5%SR
HMMT 202538.8%SR

Reasoning

HumanEvalOpenAI (2021)93.3%SR
AutoLogi89.5%SR
ZebraLogic89.0%SR
HumanEval-ER81.1%SR
EvalPlus0.80 / 100SR
MuSR76.4%SR
GPQANYU + Cohere + Anthropic (2023)75.1%SR
SWE-bench Verified (Multiple Attempts)71.6%SR
SWE-bench Verified (Agentic Coding)65.8%SR
SWE-Bench Verified65.8%SR
SuperGPQA57.2%SR
LiveCodeBench v653.7%SR
LiveCodeBench53.7%SR
SWE-bench Multilingual47.3%SR
Terminal-Bench30.0%SR
OJBench27.1%SR
Terminus25.0%SR
Humanity's Last Exam4.7%SR

AA Evaluation Indices

(Artificial Analysis)

No AA evaluation data available

LLM Stats Category Scores

(LLM Stats (zeroeval))
Instruction Following
90
Language
90
Structured Output
90
Legal
80
Finance
80
Healthcare
80
Biology
80
Chat
70
Math
70
Physics
70
Frontend Development
70
Chemistry
70
Reasoning
60
General
60
Communication
60
Economics
60
Tool Calling
60
Code
50
Factuality
30
Agents
20
Vision
0

Pricing

Input Price$0.57 / 1M tokens
Output Price$2.3 / 1M tokens
Blended Price (3:1)$1.075 / 1M tokens

Speed

Tokens/sec49.1
Time to First Token1.92s
Time to Answer1.92s

Provider Price Ranking

Provider Price Ranking

8 providers

Cheapest: OpenCodeMost Expensive: AI/ML API
ProviderInputOutput
1OpenCodeCheapest
$0.4
$2.5
2KimiPRIMARY
$0.57
$2.3
3OpenRouter
$0.57
$2.3
4Kilo Gateway
$0.57
$2.3
5Vercel AI Gateway
$0.57
$2.3
6DevPass (LLM Gateway)
$0.57
$2.3
7Qiniu
$0.57971
$2.31884
8AI/ML API
$0.60306
$2.4334

Compare pricing across different API providers for this model.

External Sources