Noometry

Model comparison

GPT-5.4 mini vs Qwen2-72B

GPT-5.4 mini is the stronger model overall, scoring 44.8 to 29.5 on the Noometry Index.

Last verified . 20 shared benchmarks.

GPT-5.4 mini OpenAI

44.8

Rank #69 Confirmed

Qwen2-72B Alibaba (Qwen)

29.5

Rank #294 Confirmed

Summary

  • They share 20 benchmarks with published results for both. GPT-5.4 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 GPT-5.4 mini leads 51.4 to 21.2.
  • The biggest single-benchmark swing is WeirdML: 60.3% for GPT-5.4 mini and 11.3% for Qwen2-72B.
  • Qwen2-72B has downloadable open weights; the other is API-only.

Side by side

GPT-5.4 miniQwen2-72B
ProviderOpenAIAlibaba (Qwen)
Noometry Index44.829.5
Released2026-03-172024-06-07
WeightsProprietaryOpen
Context window400K—
Max output128K—
Input $ / M tokens$0.75—
Output $ / M tokens$4.50—
Results tracked4626

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

Category by category

Coding GPT-5.4 mini leads

GPT-5.4 mini: 45.0 (#67), Qwen2-72B: 28.7 (#303)

BenchmarkGPT-5.4 miniQwen2-72B
WeirdML60.3%11.3%
LMArena Coding14381196
FrontierCode27%—
LMArena WebDev1397—
SciCode49.9%—
BigCodeBench Instruct—38.5%
BigCodeBench Complete—54%
ALE-Bench1,189—

Agentic & Tool Use GPT-5.4 mini leads

GPT-5.4 mini: 29.6 (#74), Qwen2-72B: 16.9 (#138)

BenchmarkGPT-5.4 miniQwen2-72B
TheAgentCompany—1.1%
DeepResearch Bench36.3%—
METR Time Horizons—29.9%

Reasoning GPT-5.4 mini leads

GPT-5.4 mini: 30.3 (#77), Qwen2-72B: 22.2 (#184)

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Hard Prompts14241191
Epoch Capabilities Index148.83125.27
ARC-AGI-218.9%—
Kagi LLM Benchmark37.9%—
NYT Connections (extended)61.8%—
ARC-AGI-163.7%—
CritPt10%—
Chess Puzzles24%—
Thematic Generalization61.7%—
Mystery Game Puzzles11%—
DTBench80%—
LMCA40.8%—
ForecastBench57—

Math GPT-5.4 mini leads

GPT-5.4 mini: 45.4 (#63), Qwen2-72B: 27.9 (#232)

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Math14181235
FrontierMath (Tiers 1-3)51.2%—
FrontierMath Tier 49.8%—
OTIS Mock AIME 2024-202588.9%—
ProofBench21%—
MATH Level 5—39.1%
FrontierMath (Feb 2025 set)28.3%—
FrontierMath Tier 4 (v1)2.1%—

Knowledge GPT-5.4 mini leads

GPT-5.4 mini: 51.4 (#62), Qwen2-72B: 21.2 (#270)

BenchmarkGPT-5.4 miniQwen2-72B
GPQA Diamond86.9%40.8%
LMArena Expert14341171
SimpleQA Verified29.4%—
Vectara Hallucination Rate5.5%—
MMLU—82.4%

Multimodal Not comparable

GPT-5.4 mini: 39.0 (#51), Qwen2-72B: —

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Vision1245—

Multilingual GPT-5.4 mini leads

GPT-5.4 mini: 52.1 (#94), Qwen2-72B: 36.0 (#239)

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Non-English14051176
LMArena Chinese14461240
LMArena French14401169
LMArena German14071151
LMArena Japanese13751111
LMArena Korean13691083
LMArena Russian14171169
LMArena Spanish14031168

Instruction Following GPT-5.4 mini leads

GPT-5.4 mini: 74.3 (#99), Qwen2-72B: 62.0 (#236)

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Instruction Following14051181

Long Context GPT-5.4 mini leads

GPT-5.4 mini: 42.9 (#107), Qwen2-72B: 36.1 (#230)

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Longer Query14071193

Writing & Preference GPT-5.4 mini leads

GPT-5.4 mini: 64.2 (#56), Qwen2-72B: 41.0 (#236)

BenchmarkGPT-5.4 miniQwen2-72B
LMArena Text14111203
LMArena Creative Writing13701181
LMArena Multi-Turn14291195
EQ-Bench Creative Writing1665—

Frequently asked questions

Is GPT-5.4 mini better than Qwen2-72B?

GPT-5.4 mini is the stronger model overall, scoring 44.8 to 29.5 on the Noometry Index.

Is GPT-5.4 mini or Qwen2-72B better for coding?

GPT-5.4 mini scores higher on coding benchmarks: 45.0 versus 28.7 in the Noometry coding category.

How many benchmarks do GPT-5.4 mini and Qwen2-72B share?

20 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and Qwen2-72B has 26.

Related comparisons

Go deeper