Noometry

Model comparison

GPT-4 vs Qwen3 32B

Qwen3 32B is the stronger model overall, scoring 38.8 to 28.9 on the Noometry Index.

Last verified . 19 shared benchmarks.

GPT-4 OpenAI

28.9

Rank #308 Confirmed

Qwen3 32B Alibaba (Qwen)

38.8

Rank #166 Confirmed

Summary

  • They share 19 benchmarks with published results for both. GPT-4 scores higher in 0 categories and Qwen3 32B in 8 categories; 8 gaps are clear of the uncertainty.
  • The widest gap is in math, where Qwen3 32B leads 37.9 to 10.1.
  • The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.1% for GPT-4 and 66.9% for Qwen3 32B.
  • Qwen3 32B is cheaper at $0.70 / $2.80 per million input/output tokens, against $30 / $60 for GPT-4.
  • Qwen3 32B accepts more context: 131K tokens versus 8K.
  • Qwen3 32B has downloadable open weights; the other is API-only.

Side by side

GPT-4Qwen3 32B
ProviderOpenAIAlibaba (Qwen)
Noometry Index28.938.8
Released2023-03-142025-04
WeightsProprietaryOpen
Context window8K131K
Max output8K16K
Input $ / M tokens$30$0.70
Output $ / M tokens$60$2.80
Results tracked3826

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

Category by category

Coding Qwen3 32B leads

GPT-4: 31.2 (#277), Qwen3 32B: 37.3 (#184)

BenchmarkGPT-4Qwen3 32B
LMArena Coding12541358
Aider Polyglot—40%
SciCode—35.4%
WeirdML12.4%—
BigCodeBench Instruct46%—
BigCodeBench Complete57.2%—
HumanEval+79.3%—

Agentic & Tool Use Not comparable

GPT-4: —, Qwen3 32B: 32.3 (#58)

BenchmarkGPT-4Qwen3 32B
Berkeley Function Calling Leaderboard—48.7%
METR Time Horizons36.1%—

Reasoning Qwen3 32B leads

GPT-4: 17.3 (#285), Qwen3 32B: 19.6 (#233)

BenchmarkGPT-4Qwen3 32B
Chess Puzzles4%5%
LMArena Hard Prompts12411335
DTBench62.7%67.5%
LMCA17.1%17.3%
Epoch Capabilities Index125.88138.51
Kagi LLM Benchmark—54.9%
CritPt—0.3%
Mystery Game Puzzles12%—
BIG-Bench Hard75.1%—
ForecastBench57.8—
HellaSwag95.3%—
WinoGrande87.5%—

Math Qwen3 32B leads

GPT-4: 10.1 (#304), Qwen3 32B: 37.9 (#89)

BenchmarkGPT-4Qwen3 32B
OTIS Mock AIME 2024-20251.1%66.9%
LMArena Math12691399
MATH Level 523%—
GSM8K92%—

Knowledge Qwen3 32B leads

GPT-4: 18.4 (#277), Qwen3 32B: 40.0 (#112)

BenchmarkGPT-4Qwen3 32B
GPQA Diamond35.7%65.7%
LMArena Expert12111362
Vectara Hallucination Rate—5.9%
MMLU86.4%—
TriviaQA84.8%—

Multilingual Qwen3 32B leads

GPT-4: 40.8 (#210), Qwen3 32B: 45.8 (#162)

BenchmarkGPT-4Qwen3 32B
LMArena Non-English12461317
LMArena Chinese12421357
LMArena German12511341
LMArena Russian12511311
LMArena French1282—
LMArena Japanese1209—
LMArena Korean1184—
LMArena Spanish1261—

Instruction Following Qwen3 32B leads

GPT-4: 65.5 (#217), Qwen3 32B: 69.1 (#174)

BenchmarkGPT-4Qwen3 32B
LMArena Instruction Following12411305

Long Context Qwen3 32B leads

GPT-4: 37.7 (#207), Qwen3 32B: 43.7 (#83)

BenchmarkGPT-4Qwen3 32B
LMArena Longer Query12441327
Fiction.LiveBench—74.2%

Writing & Preference Qwen3 32B leads

GPT-4: 35.0 (#263), Qwen3 32B: 53.1 (#158)

BenchmarkGPT-4Qwen3 32B
LMArena Text12631340
LMArena Creative Writing12441297
LMArena Multi-Turn12571331
EQ-Bench Creative Writing752—

Frequently asked questions

Is GPT-4 better than Qwen3 32B?

Qwen3 32B is the stronger model overall, scoring 38.8 to 28.9 on the Noometry Index.

Which is cheaper, GPT-4 or Qwen3 32B?

Qwen3 32B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; GPT-4 lists at $30 and $60.

Is GPT-4 or Qwen3 32B better for coding?

Qwen3 32B scores higher on coding benchmarks: 37.3 versus 31.2 in the Noometry coding category.

Which has the bigger context window?

Qwen3 32B does, with 131K tokens against 8K.

How many benchmarks do GPT-4 and Qwen3 32B share?

19 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and Qwen3 32B has 26.

Related comparisons

Go deeper