Noometry

Model comparison

Grok-3 mini vs Kimi K3

Kimi K3 is the stronger model overall, scoring 59.5 to 41.2 on the Noometry Index.

Last verified . 23 shared benchmarks.

Grok-3 mini xAI

41.2

Rank #141 Confirmed

Kimi K3 Moonshot AI

59.5

Rank #15 Confirmed

Summary

  • They share 23 benchmarks with published results for both. Grok-3 mini scores higher in 1 category and Kimi K3 in 7 categories; 7 gaps are clear of the uncertainty.
  • The widest gap is in reasoning, where Kimi K3 leads 63.0 to 13.6.
  • The biggest single-benchmark swing is ARC-AGI-1: 16.5% for Grok-3 mini and 94.5% for Kimi K3.
  • Kimi K3 has downloadable open weights; the other is API-only.

Side by side

Grok-3 miniKimi K3
ProviderxAIMoonshot AI
Noometry Index41.259.5
Released2025-04-092026-07-16
WeightsProprietaryOpen
Context window—1.05M
Max output—1.05M
Input $ / M tokens—$3
Output $ / M tokens—$15
Results tracked3553

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Kimi K3 leads

Grok-3 mini: 40.8 (#131), Kimi K3: 61.0 (#10)

BenchmarkGrok-3 miniKimi K3
WeirdML42.6%82.6%
LMArena Coding13791508
DeepSWE—68.5%
FrontierCode—44.2%
Aider Polyglot49.3%—
LMArena WebDev—1654
FrontierSWE—25.9%
SciCode—59.5%
ALE-Bench—1,524

Agentic & Tool Use Not comparable

Grok-3 mini: —, Kimi K3: 41.8 (#20)

BenchmarkGrok-3 miniKimi K3
APEX-Agents—50.6%
τ²-bench Banking—37.1%
PostTrainBench—32%
GBAEval—48.3%
GDP.pdf—19%
Vending-Bench 2—5,165

Reasoning Kimi K3 leads

Grok-3 mini: 13.6 (#334), Kimi K3: 63.0 (#17)

BenchmarkGrok-3 miniKimi K3
ARC-AGI-20.4%60.4%
ARC-AGI-116.5%94.5%
LMArena Hard Prompts13751496
Epoch Capabilities Index140.35157.45
SimpleBench—60.7%
Kagi LLM Benchmark61.3%—
NYT Connections (extended)—93.6%
CritPt—23.4%
Chess Puzzles—39%
Mystery Game Puzzles—26%
DTBench—91.2%
LMCA—52.7%
Surface Evolver Bench—95%
ForecastBench—61.1

Math Kimi K3 leads

Grok-3 mini: 42.1 (#85), Kimi K3: 74.2 (#16)

BenchmarkGrok-3 miniKimi K3
OTIS Mock AIME 2024-202577.8%97.2%
LMArena Math13861491
FrontierMath (Tiers 1-3)—72.2%
FrontierMath Tier 4—39%
MathArena Final-Answer Competitions—87.8%
ProofBench—87%
Omni-MATH31.8%—
MATH Level 590.9%—
FrontierMath (Feb 2025 set)5.9%—

Knowledge Kimi K3 leads

Grok-3 mini: 46.4 (#81), Kimi K3: 63.2 (#21)

BenchmarkGrok-3 miniKimi K3
GPQA Diamond76.3%93.1%
LMArena Expert13951521
SimpleQA Verified—50.6%
MMLU-Pro79.9%—
Confabulations10.8%—
GPQA (HELM)67.5%—

Multimodal Not comparable

Grok-3 mini: —, Kimi K3: 37.8 (#70)

BenchmarkGrok-3 miniKimi K3
Blueprint-Bench 2—29.5%
Furniture Assembly—34.2%

Multilingual Kimi K3 leads

Grok-3 mini: 48.1 (#145), Kimi K3: 56.3 (#21)

BenchmarkGrok-3 miniKimi K3
LMArena Non-English13521466
LMArena Chinese13871529
LMArena French13571491
LMArena German13491488
LMArena Japanese13421487
LMArena Korean13351458
LMArena Russian13531482
LMArena Spanish13811472

Instruction Following Too close to call

Grok-3 mini: 78.5 (#9), Kimi K3: 77.7 (#14)

BenchmarkGrok-3 miniKimi K3
LMArena Instruction Following13571483
IFEval95.1%—

Long Context Kimi K3 leads

Grok-3 mini: 41.0 (#147), Kimi K3: 45.8 (#29)

BenchmarkGrok-3 miniKimi K3
LMArena Longer Query13721494
Fiction.LiveBench66.7%—

Writing & Preference Kimi K3 leads

Grok-3 mini: 52.5 (#169), Kimi K3: 76.6 (#4)

BenchmarkGrok-3 miniKimi K3
LMArena Text13701476
LMArena Creative Writing13421454
LMArena Multi-Turn13551488
Short-Story Creative Writing73.5%—
EQ-Bench Creative Writing—2082
WildBench65.1%—
EQ-Bench 4—1339

Frequently asked questions

Is Grok-3 mini better than Kimi K3?

Kimi K3 is the stronger model overall, scoring 59.5 to 41.2 on the Noometry Index.

Is Grok-3 mini or Kimi K3 better for coding?

Kimi K3 scores higher on coding benchmarks: 61.0 versus 40.8 in the Noometry coding category.

How many benchmarks do Grok-3 mini and Kimi K3 share?

23 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Kimi K3 has 53.

Related comparisons

Go deeper