Skip to main content

Kimi K2 Thinking

KimiKimiOpen WeightMIT · Commercial OK

Description

Kimi K2 Thinking is the latest, most capable version of open-source thinking model. Starting with Kimi K2, it is built as a thinking agent that reasons step-by-step while dynamically invoking tools. It sets a new state-of-the-art on Humanity's Last Exam (HLE), BrowseComp, and other benchmarks by dramatically scaling multi-step reasoning depth and maintaining stable tool-use across 200–300 sequential calls. At the same time, K2 Thinking is a native INT4 quantization model with 256k context window, achieving lossless reductions in inference latency and GPU memory usage. Key features include deep thinking & tool orchestration with end-to-end training to interleave chain-of-thought reasoning with function calls, native INT4 quantization via Quantization-Aware Training (QAT) achieving lossless 2x speed-up, and stable long-horizon agency maintaining coherent goal-directed behavior across up to 200–300 consecutive tool invocations.

Release Date
2025-11-06
Parameters
1.0T
Context Length
262K
Modalities
text

Capability Radar

70
general
50
coding
80
reasoning
51
scienceest.
50
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))

Finance

FinSearchComp-T347.4%SR

Healthcare

HealthBench58.0%SR

Language

MMLU-Redux94.4%SR
MMLU-Pro84.6%SR

Math

AIME 2025100.0%SR
HMMT 202597.5%SR
IMO-AnswerBench78.6%SR

Reasoning

FRAMES87.0%SR
GPQANYU + Cohere + Anthropic (2023)84.5%SR
LiveCodeBench v683.1%SR
SWE-Bench Verified71.3%SR
BrowseComp-zh62.3%SR
SWE-bench Multilingual61.1%SR
BrowseCompOpenAI (2025)60.2%SR
Seal-056.3%SR
Humanity's Last Exam51.0%SR
OJBench48.7%SR
Terminal-Bench47.1%SR
SciCode44.8%SR
Multi-SWE-Bench41.9%SR

Writing

WritingBench73.8%SR

AA Evaluation Indices

(Artificial Analysis)

No AA evaluation data available

LLM Stats Category Scores

(LLM Stats (zeroeval))
Language
90
Legal
80
Math
80
Finance
80
Reasoning
70
Search
70
Frontend Development
70
General
70
Healthcare
70
Communication
70
Creativity
70
Writing
70
Physics
60
Biology
60
Chemistry
60
Agents
50
Code
50
Vision
50

Pricing

Input Price$0.6 / 1M tokens
Output Price$2.5 / 1M tokens
Blended Price (3:1)$1.075 / 1M tokens
Cache Read Price$0.15 / 1M tokens

Speed

Tokens/sec122.3
Time to First Token1.53s
Time to Answer17.87s

Provider Price Ranking

Provider Price Ranking

11 providers

Cheapest: OpenCodeMost Expensive: AI/ML API
ProviderInputOutput
1OpenCodeCheapest
$0.4
$2.5
2Helicone
$0.48
$2
3Alibaba (China)
$0.574
$2.294
4302.AI
$0.575
$2.3
5Qiniu
$0.57971
$2.31884
6KimiPRIMARY
$0.6
$2.5
7NanoGPT
$0.6
$2.5
8OpenRouter
$0.6
$2.5
9Novita AI
$0.6
$2.5
10Kilo Gateway
$0.6
$2.5
11AI/ML API
$0.6348
$2.645

Compare pricing across different API providers for this model.

External Sources