Noometry

Model comparison

o4-mini vs Qwen2-72B

o4-mini is the stronger model overall, scoring 41.6 to 30.0 on the Noometry Index.

Last verified . 22 shared benchmarks.

o4-mini OpenAI

41.6

Rank #132 Confirmed

Qwen2-72B Alibaba (Qwen)

30.0

Rank #300 Confirmed

Summary

  • They share 22 benchmarks with published results for both. o4-mini scores higher in 9 categories and Qwen2-72B in 0 categories; 9 gaps are clear of the uncertainty.
  • The widest gap is in knowledge, where o4-mini leads 43.6 to 21.2.
  • The biggest single-benchmark swing is MATH Level 5: 97.8% for o4-mini and 39.1% for Qwen2-72B.
  • Qwen2-72B has downloadable open weights; the other is API-only.

Side by side

o4-miniQwen2-72B
ProviderOpenAIAlibaba (Qwen)
Noometry Index41.630.0
Released2025-04-162024-06-07
WeightsProprietaryOpen
Context window200K—
Max output100K—
Input $ / M tokens$1.10—
Output $ / M tokens$4.40—
Results tracked6026

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

Category by category

Coding o4-mini leads

o4-mini: 40.9 (#127), Qwen2-72B: 29.1 (#310)

Benchmarko4-miniQwen2-72B
WeirdML52.6%11.3%
LMArena Coding13681196
SWE-bench Verified (bash only)45%—
Aider Polyglot72%—
GSO3.6%—
BigCodeBench Instruct—38.5%
BigCodeBench Complete—54%
CadEval62%—
ALE-Bench826.17—
AlgoTune1.72—

Agentic & Tool Use o4-mini leads

o4-mini: 32.6 (#61), Qwen2-72B: 17.0 (#146)

Benchmarko4-miniQwen2-72B
METR Time Horizons63.9%29.9%
Berkeley Function Calling Leaderboard53.2%—
GDPval25.3%—
TheAgentCompany—1.1%

Reasoning o4-mini leads

o4-mini: 24.6 (#162), Qwen2-72B: 23.2 (#181)

Benchmarko4-miniQwen2-72B
LMArena Hard Prompts13511191
Epoch Capabilities Index145.64125.28
ARC-AGI-26.1%—
SimpleBench38.7%—
Kagi LLM Benchmark67.6%—
ARC-AGI-158.7%—
CritPt0.6%—
Chess Puzzles26%—
EnigmaEval9.2%—
Mystery Game Puzzles5%—
DTBench77.6%—
LMCA26.5%—
ForecastBench61.8—

Math o4-mini leads

o4-mini: 40.8 (#89), Qwen2-72B: 30.2 (#236)

Benchmarko4-miniQwen2-72B
LMArena Math13891235
MATH Level 597.8%39.1%
FrontierMath (Tiers 1-3)36.1%—
FrontierMath Tier 44.9%—
OTIS Mock AIME 2024-202581.7%—
Omni-MATH72%—
FrontierMath (Feb 2025 set)24.8%—
FrontierMath Tier 4 (v1)6.3%—

Knowledge o4-mini leads

o4-mini: 43.6 (#91), Qwen2-72B: 21.2 (#275)

Benchmarko4-miniQwen2-72B
GPQA Diamond79.6%40.8%
LMArena Expert13431171
Humanity's Last Exam18.1%—
SimpleQA Verified19.6%—
MMLU-Pro82%—
Confabulations15.8%—
Vectara Hallucination Rate18.6%—
GPQA (HELM)73.5%—
MMLU—82.4%

Multimodal Not comparable

o4-mini: 40.2 (#49), Qwen2-72B: —

Benchmarko4-miniQwen2-72B
LMArena Vision1194—
GeoBench64%—
VPCT57.5%—

Multilingual o4-mini leads

o4-mini: 47.0 (#154), Qwen2-72B: 35.9 (#244)

Benchmarko4-miniQwen2-72B
LMArena Non-English13371176
LMArena Chinese13541240
LMArena French13641170
LMArena German13361151
LMArena Japanese13081111
LMArena Korean13121083
LMArena Russian13341169
LMArena Spanish13471169

Instruction Following o4-mini leads

o4-mini: 75.2 (#68), Qwen2-72B: 61.7 (#241)

Benchmarko4-miniQwen2-72B
LMArena Instruction Following13211181
IFEval92.8%—

Long Context o4-mini leads

o4-mini: 45.5 (#33), Qwen2-72B: 36.1 (#235)

Benchmarko4-miniQwen2-72B
LMArena Longer Query13151192
Fiction.LiveBench77.8%—

Writing & Preference o4-mini leads

o4-mini: 54.0 (#152), Qwen2-72B: 40.8 (#241)

Benchmarko4-miniQwen2-72B
LMArena Text13531203
LMArena Creative Writing12941181
LMArena Multi-Turn13501196
Short-Story Creative Writing75%—
WildBench85.4%—

Frequently asked questions

Is o4-mini better than Qwen2-72B?

o4-mini is the stronger model overall, scoring 41.6 to 30.0 on the Noometry Index.

Is o4-mini or Qwen2-72B better for coding?

o4-mini scores higher on coding benchmarks: 40.9 versus 29.1 in the Noometry coding category.

How many benchmarks do o4-mini and Qwen2-72B share?

22 benchmarks have published results for both models. o4-mini has 60 scored results on Noometry and Qwen2-72B has 26.

Related comparisons

Go deeper