Kimi K2 Jul 2025

Kimi K2 ranks first on 3 of the 77 benchmarks it has scores on.

Moonshot AI Released 12 Jul 2025 Family Kimi K2 The lab's model card

Benchmarks with a score

Ranked first, out of five or more models

In the top three

Moonshot AI Jul 2025

Kimi K2

Jul 2025 snapshot Open weights Flagship

139.9 #123 of 271

Epoch Capabilities Index

SWE-Bench Pro 27.7% #17 /24
LongBench v2 49.1% #20 /27
SuperGPQA 57.2% #20 /26

Where it ranks

Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.

Knowledge and language

Benchmark	Score	Rank	Source
DROP	93.5%	#1 of 40	Model card
Lech Mazur Creative Writing	8.56	#2 of 44	Epoch AI
WritingBench	81.3	#3 of 45	Official leaderboard
FACTS Grounding	88.5%	#1 of 15	Model card
EQ-Bench 3	1,548	#11 of 78	Official leaderboard
MMLU	89.5%	#19 of 127	Model card
IFEval	89.8%	#26 of 126	Model card
Creative Writing v3	75.0	#30 of 120	Official leaderboard
LMArena (Elo)	1,418	#106 of 333	Official leaderboard
SimpleQA	31%	#22 of 59	Model card
Longform Creative Writing (EQ-Bench)	68.1	#48 of 119	Official leaderboard
MMLU-Redux	92.7%	#16 of 38	Model card
MMLU-Pro	81.1%	#67 of 151	Model card
OpenAI MRCR	55%	#5 of 10	Model card
Judgemark v2.1	73.1	#22 of 42	Official leaderboard
Fiction.LiveBench	61.1%	#34 of 57	Epoch AI
WideSearch	3.00%	#4 of 6	Official leaderboard
LLM Confabulation Leaderboard	20.4%	#39 of 58	Official leaderboard
MultiChallenge	54.1%	#25 of 36	Model card
LongBench v2	49.1%	#20 of 27	Model card
FRAMES	77.1%	#6 of 8	Model card
MultiNRC	18.5%	#29 of 38	Official leaderboard
SuperGPQA	57.2%	#20 of 26	Model card
Arena-Hard v2.0	54.5%	#19 of 24	Model card
Vectara Hallucination Leaderboard (HHEM)	1.10%	#145 of 148	Official leaderboard
Humanity's Last Exam	4.70%	#146 of 147	Model card
FaithJudge	7.40%	#1 of 1	Model card

Math and science

Benchmark	Score	Rank	Source
AutoLogi	89.5%	#1 of 8	Model card
MATH-500	97.4%	#14 of 67	Model card
ZebraLogic	89%	#8 of 23	Model card
PolyMath	65.1%	#7 of 18	Model card
ForecastBench	62.1	#37 of 86	Official leaderboard
GPQA Diamond	75.1%	#140 of 284	Model card
HiPhO	54.4%	#19 of 35	Official leaderboard
PHYBench	21.3%	#16 of 28	Official leaderboard
AIME 2024	69.6%	#66 of 87	Model card
CNMO 2024	74.3%	#5 of 6	Model card
SimpleBench	26.3%	#84 of 100	Official leaderboard
AIME 2025	49.5%	#132 of 143	Model card
HMMT 2025	38.8%	#49 of 49	Model card

Coding

Benchmark	Score	Rank	Source
MultiPL-E	85.7%	#2 of 18	Model card
AutoCodeBench	47.8%	#10 of 34	Official leaderboard
Aider Polyglot	59.1%	#19 of 54	Epoch AI
SWE-bench Verified	65.4%	#22 of 60	Official leaderboard
PaperBench	27.8%	#8 of 14	Model card
ArtifactsBench	47.6%	#13 of 20	Official leaderboard
SWE-Bench Pro	27.7%	#17 of 24	Official leaderboard
WeirdML	39.4%	#81 of 112	Official leaderboard
GSO-Bench	4.90%	#22 of 30	Official leaderboard
LiveCodeBench v6	53.7%	#67 of 76	Model card
Multi-SWE-bench	18.3%	#9 of 10	Model card
SWE-Lancer	39.1%	#9 of 10	Model card
OJBench	27.1%	#11 of 12	Model card
SWE-bench Multilingual	47.3%	#66 of 70	Model card
IOI (Vals)	1.25%	#52 of 55	Official leaderboard

Agents and tools

Benchmark	Score	Rank	Source
AceBench	82.2%	#2 of 16	Official leaderboard
BFCL V4	59.1%	#9 of 65	Official leaderboard
Galileo Agent Leaderboard v2	53%	#5 of 21	Official leaderboard
Terminal-Bench	27.8%	#9 of 18	Epoch AI
LisanBench	347.3	#77 of 133	Official leaderboard
MCPMark	19.1%	#17 of 28	Official leaderboard
Elimination Game	3	#40 of 56	Official leaderboard
Terminal-Bench 2.0	27.8%	#32 of 43	Epoch AI
tau2-bench	66.1%	#57 of 69	Model card
Step Race	1	#49 of 59	Official leaderboard
MCP-Universe	19.1%	#27 of 30	Official leaderboard
LLM Chess	-689	#120 of 124	Official leaderboard

Professional work

Benchmark	Score	Rank	Source
LegalBench	81.5%	#78 of 138	Official leaderboard
TaxEval v2	70.2%	#86 of 129	Official leaderboard
CorpFin v2	50.4%	#97 of 119	Official leaderboard
PRBench	36.4%	#41 of 44	Official leaderboard

Safety and security

Benchmark	Score	Rank	Source
HELM Safety	98.0	#7 of 75	Official leaderboard
SimpleSafetyTests	100%	#16 of 75	Official leaderboard
AIR-Bench 2024	74.1%	#30 of 82	Official leaderboard
Anthropic Red Team	99.3%	#30 of 75	Official leaderboard
FORTRESS	55.5%	#57 of 66	Official leaderboard

Other

Benchmark	Score		Rank	Source
HELM Capabilities	76.8		#10 of 60	Official leaderboard
LiveBench	76.4%		#7 of 27	Model card

Moonshot AI's models over time

Moonshot AI's models in color, against every other model on the Epoch Capabilities Index, by release date.

Its launch post

The benchmarks the lab chose to headline, with the scores it gave.

Kimi K2 Instruct model card (Hugging Face) 11 Jul 2025

Its system card

Every result in the longer documents the lab published with it. 14 of them aren't in the launch post.

Kimi K2: Open Agentic Intelligence (arXiv 2507.20534) 21 Jul 2025