ModelsIndexCompareBest forBenchmarksStatusPricingAPI

Kimi K2

Moonshot AI·Operational·Released 2025-07-11text
Context128k
Max output-
Input / 1M-
Output / 1M-
Cached in-
Benchmarks18

Benchmarks

Cited public sources · rank is exact-benchmark, apples-to-apples

SimpleBench (SimpleBench)#67/7326.3% ↗
WeirdML (Håvard Tveit Ihle)#71/9239.4% ↗
GSO Opt@1 (GSO)#24/272% ↗
Aider-Polyglot#3/860% ↗
AIME 2025#24/2449.5% ↗
GPQA-Diamond#6/875.1% ↗
Humanity's Last Exam (Text Only)4.7% ↗
IFEval (Prompt Strict)89.8% ↗
LiveCodeBench v6 (Aug 24 - May 25)53.7% ↗
MMLU-Pro#28/4981.1% ↗
MMLU-Redux#4/1392.7% ↗
SimpleQA#6/731% ↗
SWE-bench Multilingual (Agentic Coding, single attempt)47.3% ↗
SWE-bench Verified (Agentic Coding, single attempt)65.8% ↗
Tau2 airline56.5% ↗
Tau2 retail70.6% ↗
Tau2 telecom65.8% ↗
TerminalBench (Inhouse Framework)30% ↗

Compare

vs Gemini 4 Argonvs GPT-6.1 Solvs Claude Sonnet 5.5vs Muse Spark 1.3vs GPT-6 Solvs GPT-6 Lunavs Claude Opus 5.5vs GPT-6 Astra

More from Moonshot AI

Kimi K3Kimi K2.7 CodeKimi K2.6Kimi K2.5Kimi K2 ThinkingKimi K2-0905

Kimi K2 FAQ

What is Kimi K2?
Kimi K2 is a large language model from Moonshot AI, released on July 11, 2025. It accepts text input.
When was Kimi K2 released?
Moonshot AI released Kimi K2 on July 11, 2025.
What is Kimi K2's context window?
Kimi K2 has a 128K-token context window.
How does Kimi K2 score on benchmarks?
Kimi K2 has 18 cited benchmark scores, including SimpleBench (SimpleBench) 26.3%, WeirdML (Håvard Tveit Ihle) 39.4% and GSO Opt@1 (GSO) 2%. Each score links to its source.

Access Kimi K2 and every other model through one endpoint with automatic failover: Respan gateway.