Noometry

Model comparison

GPT-5.4 nano vs Qwen3 32B

GPT-5.4 nano is the stronger model overall, scoring 41.9 to 39.2 on the Noometry Index.

Last verified . 23 shared benchmarks.

GPT-5.4 nano OpenAI

41.9

Rank #125 Confirmed

Qwen3 32B Alibaba (Qwen)

39.2

Rank #172 Confirmed

Summary

  • They share 23 benchmarks with published results for both. GPT-5.4 nano scores higher in 7 categories and Qwen3 32B in 1 category; 8 gaps are clear of the uncertainty.
  • The widest gap is in coding, where GPT-5.4 nano leads 43.6 to 37.7.
  • The biggest single-benchmark swing is Chess Puzzles: 30% for GPT-5.4 nano and 5% for Qwen3 32B.
  • GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $0.70 / $2.80 for Qwen3 32B.
  • GPT-5.4 nano accepts more context: 400K tokens versus 131K.
  • Qwen3 32B has downloadable open weights; the other is API-only.

Side by side

GPT-5.4 nanoQwen3 32B
ProviderOpenAIAlibaba (Qwen)
Noometry Index41.939.2
Released2026-03-172025-04
WeightsProprietaryOpen
Context window400K131K
Max output128K16K
Input $ / M tokens$0.20$0.70
Output $ / M tokens$1.25$2.80
Results tracked4026

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

Category by category

Coding GPT-5.4 nano leads

GPT-5.4 nano: 43.6 (#84), Qwen3 32B: 37.7 (#190)

BenchmarkGPT-5.4 nanoQwen3 32B
SciCode46.9%35.4%
LMArena Coding14051358
Aider Polyglot—40%
WeirdML49.2%—
ALE-Bench1,005—

Agentic & Tool Use Not comparable

GPT-5.4 nano: —, Qwen3 32B: 32.6 (#62)

BenchmarkGPT-5.4 nanoQwen3 32B
Berkeley Function Calling Leaderboard—48.7%

Reasoning GPT-5.4 nano leads

GPT-5.4 nano: 23.7 (#173), Qwen3 32B: 20.2 (#241)

BenchmarkGPT-5.4 nanoQwen3 32B
Kagi LLM Benchmark39.7%54.9%
CritPt9.3%0.3%
Chess Puzzles30%5%
LMArena Hard Prompts13811334
DTBench80.3%67.5%
LMCA36.9%17.3%
Epoch Capabilities Index145.81138.51
ARC-AGI-25.7%—
ARC-AGI-151.5%—
Mystery Game Puzzles9%—
ForecastBench57.3—

Math GPT-5.4 nano leads

GPT-5.4 nano: 40.9 (#88), Qwen3 32B: 39.7 (#99)

BenchmarkGPT-5.4 nanoQwen3 32B
OTIS Mock AIME 2024-202587.8%66.9%
LMArena Math14061399
FrontierMath (Tiers 1-3)44.9%—
FrontierMath Tier 412.2%—
ProofBench5%—
FrontierMath (Feb 2025 set)25.9%—
FrontierMath Tier 4 (v1)6.3%—

Knowledge GPT-5.4 nano leads

GPT-5.4 nano: 41.9 (#103), Qwen3 32B: 40.0 (#125)

BenchmarkGPT-5.4 nanoQwen3 32B
GPQA Diamond78.5%65.7%
Vectara Hallucination Rate3.1%5.9%
LMArena Expert13961362
SimpleQA Verified11.7%—

Multimodal Not comparable

GPT-5.4 nano: 36.7 (#78), Qwen3 32B: —

BenchmarkGPT-5.4 nanoQwen3 32B
LMArena Vision1196—

Multilingual GPT-5.4 nano leads

GPT-5.4 nano: 48.6 (#140), Qwen3 32B: 45.6 (#167)

BenchmarkGPT-5.4 nanoQwen3 32B
LMArena Non-English13591317
LMArena Chinese13921357
LMArena German13671341
LMArena Russian13631311
LMArena French1396—
LMArena Japanese1343—
LMArena Korean1320—
LMArena Spanish1371—

Instruction Following GPT-5.4 nano leads

GPT-5.4 nano: 71.9 (#144), Qwen3 32B: 68.9 (#179)

BenchmarkGPT-5.4 nanoQwen3 32B
LMArena Instruction Following13621305

Long Context Qwen3 32B leads

GPT-5.4 nano: 41.6 (#137), Qwen3 32B: 43.8 (#87)

BenchmarkGPT-5.4 nanoQwen3 32B
LMArena Longer Query13661327
Fiction.LiveBench—74.2%

Writing & Preference GPT-5.4 nano leads

GPT-5.4 nano: 55.7 (#142), Qwen3 32B: 52.9 (#163)

BenchmarkGPT-5.4 nanoQwen3 32B
LMArena Text13721340
LMArena Creative Writing13141297
LMArena Multi-Turn13821331

Frequently asked questions

Is GPT-5.4 nano better than Qwen3 32B?

GPT-5.4 nano is the stronger model overall, scoring 41.9 to 39.2 on the Noometry Index.

Which is cheaper, GPT-5.4 nano or Qwen3 32B?

GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; Qwen3 32B lists at $0.70 and $2.80.

Is GPT-5.4 nano or Qwen3 32B better for coding?

GPT-5.4 nano scores higher on coding benchmarks: 43.6 versus 37.7 in the Noometry coding category.

Which has the bigger context window?

GPT-5.4 nano does, with 400K tokens against 131K.

How many benchmarks do GPT-5.4 nano and Qwen3 32B share?

23 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Qwen3 32B has 26.

Related comparisons

Go deeper