Noometry

Model comparison

Grok-3 mini vs Qwen2.5 72B Instruct

Grok-3 mini is the stronger model overall, scoring 41.0 to 31.5 on the Noometry Index.

Last verified . 28 shared benchmarks.

Grok-3 mini xAI

41.0

Rank #130 Confirmed

Qwen2.5 72B Instruct Alibaba (Qwen)

31.5

Rank #259 Confirmed

Summary

  • They share 28 benchmarks with published results for both. Grok-3 mini scores higher in 7 categories and Qwen2.5 72B Instruct in 1 category; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Grok-3 mini leads 40.3 to 18.1.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 77.8% for Grok-3 mini and 8.1% for Qwen2.5 72B Instruct.
  • Qwen2.5 72B Instruct has downloadable open weights; the other is API-only.

Side by side

Grok-3 miniQwen2.5 72B Instruct
ProviderxAIAlibaba (Qwen)
Noometry Index41.031.5
Released2025-04-092024-09
WeightsProprietaryOpen
Context window—131K
Max output—8K
Input $ / M tokens—$1.40
Output $ / M tokens—$5.60
Results tracked3543

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

Category by category

Coding Grok-3 mini leads

Grok-3 mini: 40.4 (#126), Qwen2.5 72B Instruct: 32.8 (#256)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
WeirdML42.6%16%
LMArena Coding13801293
Aider Polyglot49.3%—
BigCodeBench Instruct—45.8%
BigCodeBench Complete—55.9%

Agentic & Tool Use Not comparable

Grok-3 mini: —, Qwen2.5 72B Instruct: 22.0 (#124)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
TheAgentCompany—5.7%
BALROG—16.2%
METR Time Horizons—35.8%

Reasoning Qwen2.5 72B Instruct leads

Grok-3 mini: 13.8 (#323), Qwen2.5 72B Instruct: 21.4 (#197)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
LMArena Hard Prompts13741271
Epoch Capabilities Index140.35129.01
ARC-AGI-20.4%—
Kagi LLM Benchmark61.3%—
ARC-AGI-116.5%—
DTBench—62.9%
LMCA—13.4%
BIG-Bench Hard—79.8%
ForecastBench—57.5
HellaSwag—84.8%
PIQA—82.6%
WinoGrande—82.3%

Math Grok-3 mini leads

Grok-3 mini: 40.3 (#78), Qwen2.5 72B Instruct: 18.1 (#281)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
OTIS Mock AIME 2024-202577.8%8.1%
Omni-MATH31.8%33%
LMArena Math13861283
MATH Level 590.9%63.2%
FrontierMath (Feb 2025 set)5.9%—

Knowledge Grok-3 mini leads

Grok-3 mini: 46.3 (#75), Qwen2.5 72B Instruct: 26.9 (#247)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
GPQA Diamond76.3%49.1%
MMLU-Pro79.9%63.1%
Confabulations10.8%19.1%
GPQA (HELM)67.5%42.6%
LMArena Expert13951245
ARC (AI2) Challenge—94.5%
MMLU—85.3%
TriviaQA—71.9%

Multilingual Grok-3 mini leads

Grok-3 mini: 48.3 (#141), Qwen2.5 72B Instruct: 41.2 (#208)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
LMArena Non-English13521252
LMArena Chinese13871272
LMArena French13571279
LMArena German13501234
LMArena Japanese13431180
LMArena Korean13361188
LMArena Russian13521264
LMArena Spanish13791255

Instruction Following Grok-3 mini leads

Grok-3 mini: 78.7 (#8), Qwen2.5 72B Instruct: 65.7 (#216)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
IFEval95.1%80.6%
LMArena Instruction Following13561254

Long Context Grok-3 mini leads

Grok-3 mini: 41.0 (#142), Qwen2.5 72B Instruct: 38.9 (#183)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
LMArena Longer Query13721282
Fiction.LiveBench66.7%—

Writing & Preference Grok-3 mini leads

Grok-3 mini: 52.7 (#163), Qwen2.5 72B Instruct: 46.9 (#210)

BenchmarkGrok-3 miniQwen2.5 72B Instruct
LMArena Text13701269
LMArena Creative Writing13421221
WildBench65.1%80.2%
LMArena Multi-Turn13551272
Short-Story Creative Writing73.5%—

Frequently asked questions

Is Grok-3 mini better than Qwen2.5 72B Instruct?

Grok-3 mini is the stronger model overall, scoring 41.0 to 31.5 on the Noometry Index.

Is Grok-3 mini or Qwen2.5 72B Instruct better for coding?

Grok-3 mini scores higher on coding benchmarks: 40.4 versus 32.8 in the Noometry coding category.

How many benchmarks do Grok-3 mini and Qwen2.5 72B Instruct share?

28 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Qwen2.5 72B Instruct has 43.

Related comparisons

Go deeper