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

Kimi K2 Thinking

by moonshotai · Released Nov 2025

Open Source
58.3
avg score
Rank #100
Compare
Better than 68% of all models
Context
262K tokens (~131 books)
Input $/1M
$0.60
Output $/1M
$2.50
Type
text
License
Open Source
Benchmarks
27 tested
Data as of
About

Kimi K2 Thinking is Moonshot AI’s most advanced open reasoning model to date, extending the K2 series into agentic, long-horizon reasoning. Built on the trillion-parameter Mixture-of-Experts (MoE) architecture introduced in...

Tested on 27 benchmarks · BenchGecko score 58.3. Top scores: OpenCompass — AIME2025 (94.1%), OpenCompass — IFEval (92.4%), OpenCompass — MMLU-Pro (84.3%).

Capabilities
coding
54.1
#74 globally
reasoning
57.9
#63 globally
math
55.9
#98 globally
knowledge
48.5
#135 globally
agentic
4.0
#66 globally
language
73.6
#73 globally
general
38.4
#76 globally
Benchmark Scores
Compare All
Tested on 27 benchmarks · Ranked across 7 categories
Score Distribution (all 312 models)
0255075100
▲ You are here
codingCompare coding →
OpenCompass — LiveCodeBenchV6

OpenCompass Live Code Bench v6. Fresh competitive programming problems to evaluate code generation without memorization.

77.1·
LiveBench — Coding

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

67.4·
SWE-Bench Verified (Bash Only)

SWE-bench Verified solved using only bash commands, no specialized frameworks. Tests raw terminal-based problem solving.

63.4·
reasoningCompare reasoning →
LiveBench — Reasoning

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

63.5·
LiveBench — Data Analysis

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

52.3·
mathCompare math →
OpenCompass — AIME2025

OpenCompass evaluation on AIME 2025 problems. Tests mathematical reasoning on fresh competition problems.

94.1·
OTIS Mock AIME 2024-2025

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

83.0·
LiveBench — Mathematics

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

81.1·
Quick compare:OpenAI logovs GPT-5.2Alibaba Qwen logovs Qwen3.6 Max PreviewOpenAI logovs GPT-4o (2024-05-13)
Excellent (85+) Good (70-85) Average (50-70) Below (<50)
Recently Happened
Kimi K2 Thinking pricing increased 28%
Apr 9, 2026
Similar Models
OpenAI logo
GPT-5.2
OpenAI
58.4$1.75/1M
Alibaba Qwen logo
Qwen3.6 Max Preview
Alibaba Qwen
58.4$1.03/1M
OpenAI logo
GPT-4o (2024-05-13)
OpenAI
58.5$5.00/1M
Links
Info
moonshotaiPricing explorerDevelopers · API
Research
Technical Report
Documentation
API DocsPlayground
Community
@moonshotai
Source Code
GitHubHugging Face
BenchGecko API
kimi-k2-thinking
Specifications
  • Typetext
  • Context262K tokens (~131 books)
  • ReleasedNov 2025
  • LicenseOpen Source
  • StatusActive
  • Cost / Message~$0.004
Available On
moonshotai logomoonshotai$0.60
Categories
codingreasoningmathknowledgeagenticlanguagegeneral
Learn More
context-windowtransformeropen weightstokens
Share & Export
Tweet
Related Models
OpenAI logoGPT-5.2
58.4
Alibaba Qwen logoQwen3.6 Max Preview
58.4
OpenAI logoGPT-4o (2024-05-13)
58.5
Alibaba logoQwen2.5 32B Instruct
57.9
Kimi K2 Thinking is an open-source text AI model by moonshotai, released in November 2025. It has an average benchmark score of 58.3. Context window: 262K tokens.

Related Models

GPT-5.2 · OpenAIQwen3.6 Max Preview · Alibaba QwenGPT-4o (2024-05-13) · OpenAIQwen2.5 32B Instruct · AlibabaGemini 2.5 Pro · Google DeepMind

Benchmarks

OpenCompass — AIME2025OpenCompass — IFEvalOpenCompass — MMLU-ProOTIS Mock AIME 2024-2025OpenCompass — GPQA-Diamond

Related Pages

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

Key facts · as of 2026-10-05

  • Kimi K2 Thinking by moonshotai. BenchGecko score 58.3, rank 100 of 312 scored models (normalized average of public benchmark scores).
  • List price $0.60 input · $2.50 output per 1M tokens (as of 2026-10-05).
  • Sold by 2 providers (as of 2026-10-05): Google $0.60 in / $2.50 out · Novita (bf16) $0.60 in / $2.50 out. Every provider
  • 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 Thinking · benchmarks, pricing and providers. BenchGecko, data as of 2026-10-05. https://benchgecko.ai/model/kimi-k2-thinking

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