NEWHow AI models draw the world map · Gecko Tests
Beta
Home/Models/Kimi K2.7 Code
moonshotai logo

Kimi K2.7 Code

by moonshotai · Released Jun 2026

Open SourceMultimodal
63.6
avg score
Rank #68
Compare
Better than 78% of all models
Context
262K tokens (~131 books)
Input $/1M
$0.61
Output $/1M
$3.07
Type
multimodal
License
Open Source
Benchmarks
32 tested
Data as of
About

MoonshotAI: Kimi K2.7 Code is a coding-focused model in Moonshot AI's Kimi K2 family, built to complete end-to-end programming tasks reliably over long contexts. It uses a native multimodal mixture-of-experts...

Tested on 32 benchmarks · BenchGecko score 63.6. Top scores: OTIS Mock AIME 2024-2025 (95.5%), Artificial Analysis · tau2-Bench Telecom (90.1%), Artificial Analysis · GPQA Diamond (89.6%).

Looking for similar performance at lower cost?
Qwen3 30B A3B Thinking 2507 scores 63.5 (100% as good) at $0.20/1M input · 67% cheaper
Capabilities
coding
51.7
#86 globally
reasoning
65.0
#50 globally
math
60.3
#85 globally
knowledge
52.3
#108 globally
agentic
37.6
#33 globally
language
67.1
#89 globally
general
48.8
#34 globally
speed
50.2
#32 globally
Benchmark Scores
Compare All
Tested on 32 benchmarks · Ranked across 8 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
codingCompare coding →
LiveBench — Coding

Regularly refreshed coding problems that avoid data contamination. New problems added monthly to prevent memorization.

74.0·
LiveBench — Agentic Coding

LiveBench coding tasks that require multi-step reasoning and tool use. Tests planning and execution of complex coding workflows.

70.0·
WeirdML

Unusual and adversarial machine learning challenges. Tests robustness of reasoning about edge cases in ML systems.

54.1·
reasoningCompare reasoning →
LiveBench — Reasoning

Regularly refreshed reasoning problems testing logical deduction, spatial reasoning, and analytical thinking.

82.8·
LiveBench — Data Analysis

Fresh data analysis tasks testing ability to interpret tables, charts, and statistical data.

62.7·
SimpleBench

Deceptively simple questions that humans find easy but AI models often get wrong. Tests common sense and reasoning gaps.

49.5·
mathCompare math →
OTIS Mock AIME 2024-2025

Mock AIME (American Invitational Mathematics Exam) problems from OTIS. Tests mathematical competition performance.

95.5·
LiveBench — Mathematics

Regularly updated math problems that test numerical reasoning, algebra, calculus, and combinatorics.

79.6·
FrontierMath-Tiers-1-3-v2-Private
54.0·
Quick compare:Mistral AI logovs Mixtral 8x7BAlibaba Qwen logovs Qwen3 30B A3B Thinking 2507OpenAI logovs GPT-4 (older v0314)
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Similar Models
Mistral AI logo
Mixtral 8x7B
Mistral AI
63.6n/a
Alibaba Qwen logo
Qwen3 30B A3B Thinking 2507
Alibaba Qwen
63.5$0.20/1M
OpenAI logo
GPT-4 (older v0314)
OpenAI
63.8$30.00/1M
Links
Info
moonshotaiPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@moonshotai
Source Code
GitHubHugging Face
BenchGecko API
kimi-k2-7-code
Specifications
  • Typemultimodal
  • Context262K tokens (~131 books)
  • ReleasedJun 2026
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.004
Available On
moonshotai logomoonshotai$0.61
Categories
codingreasoningmathknowledgeagenticlanguagegeneralspeed
Learn More
context-windowtransformeropen weightstokens
Share & Export
Tweet
Related Models
Mistral AI logoMixtral 8x7B
63.6
Alibaba Qwen logoQwen3 30B A3B Thinking 2507
63.5
OpenAI logoGPT-4 (older v0314)
63.8
z-ai logoGLM 5.1
63.8
Kimi K2.7 Code is an open-source multimodal AI model by moonshotai, released in June 2026. It has an average benchmark score of 63.6. Context window: 262K tokens.

Related Models

Mixtral 8x7B · Mistral AIQwen3 30B A3B Thinking 2507 · Alibaba QwenGPT-4 (older v0314) · OpenAIGLM 5.1 · z-aiGPT-5.1-Codex-Mini · OpenAI

Benchmarks

OTIS Mock AIME 2024-2025Artificial Analysis · tau2-Bench TelecomArtificial Analysis · GPQA DiamondGPQA diamondLiveBench — Reasoning

Related Pages

moonshotai · Providermoonshotai · EconomyAll ModelsCompare ModelsPricingDevelopers · APIContext Window · Glossary

Key facts · as of 2026-10-05

  • Kimi K2.7 Code by moonshotai. BenchGecko score 63.6, rank 67 of 312 scored models (normalized average of public benchmark scores).
  • List price $0.61 input · $3.07 output per 1M tokens (as of 2026-10-05).
  • Gecko Tests: Who Are You B (Knows who made it) · Tokenizer Tax E (92% more tokens outside English).

How to cite · data as of 2026-10-05

Kimi K2.7 Code · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/kimi-k2-7-code

Credit "Source: BenchGecko" with a link. Prices per provider and Gecko Tests are BenchGecko data (CC BY 4.0); benchmark scores keep their original source, listed in the JSON. JSON · llms.txt · MCP