Model comparison
GPT-5.4 nano vs Qwen2.5 32B Instruct
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 30.1 on the Noometry Index.
Last verified . 4 shared benchmarks.
Alibaba (Qwen)Summary
- They share 4 benchmarks with published results for both. GPT-5.4 nano scores higher in 4 categories and Qwen2.5 32B Instruct in 0 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.4 nano leads 40.9 to 16.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 87.8% for GPT-5.4 nano and 7.4% for Qwen2.5 32B Instruct.
- GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $0.70 / $2.80 for Qwen2.5 32B Instruct.
- GPT-5.4 nano accepts more context: 400K tokens versus 131K.
- Qwen2.5 32B Instruct has downloadable open weights; the other is API-only.
Side by side
| GPT-5.4 nano | Qwen2.5 32B Instruct | |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 41.9 | 30.1 |
| Released | 2026-03-17 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 400K | 131K |
| Max output | 128K | 8K |
| Input $ / M tokens | $0.20 | $0.70 |
| Output $ / M tokens | $1.25 | $2.80 |
| Results tracked | 40 | 7 |
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), Qwen2.5 32B Instruct: 38.7 (#169)
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| SciCode | 46.9% | — |
| WeirdML | 49.2% | — |
| BigCodeBench Instruct | — | 45% |
| LMArena Coding | 1405 | — |
| BigCodeBench Complete | — | 52.3% |
| ALE-Bench | 1,005 | — |
Reasoning GPT-5.4 nano leads
GPT-5.4 nano: 23.7 (#173), Qwen2.5 32B Instruct: 19.2 (#266)
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| Chess Puzzles | 30% | 0% |
| Epoch Capabilities Index | 145.81 | 128.52 |
| ARC-AGI-2 | 5.7% | — |
| Kagi LLM Benchmark | 39.7% | — |
| ARC-AGI-1 | 51.5% | — |
| CritPt | 9.3% | — |
| LMArena Hard Prompts | 1381 | — |
| Mystery Game Puzzles | 9% | — |
| DTBench | 80.3% | — |
| LMCA | 36.9% | — |
| ForecastBench | 57.3 | — |
Math GPT-5.4 nano leads
GPT-5.4 nano: 40.9 (#88), Qwen2.5 32B Instruct: 16.2 (#296)
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 87.8% | 7.4% |
| FrontierMath (Tiers 1-3) | 44.9% | — |
| FrontierMath Tier 4 | 12.2% | — |
| ProofBench | 5% | — |
| LMArena Math | 1406 | — |
| MATH Level 5 | — | 56.1% |
| 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), Qwen2.5 32B Instruct: 24.9 (#266)
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| GPQA Diamond | 78.5% | 46.1% |
| SimpleQA Verified | 11.7% | — |
| Vectara Hallucination Rate | 3.1% | — |
| LMArena Expert | 1396 | — |
Multimodal Not comparable
GPT-5.4 nano: 36.7 (#78), Qwen2.5 32B Instruct: —
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| LMArena Vision | 1196 | — |
Multilingual Not comparable
GPT-5.4 nano: 48.6 (#140), Qwen2.5 32B Instruct: —
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| LMArena Non-English | 1359 | — |
| LMArena Chinese | 1392 | — |
| LMArena French | 1396 | — |
| LMArena German | 1367 | — |
| LMArena Japanese | 1343 | — |
| LMArena Korean | 1320 | — |
| LMArena Russian | 1363 | — |
| LMArena Spanish | 1371 | — |
Instruction Following Not comparable
GPT-5.4 nano: 71.9 (#144), Qwen2.5 32B Instruct: —
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| LMArena Instruction Following | 1362 | — |
Long Context Not comparable
GPT-5.4 nano: 41.6 (#137), Qwen2.5 32B Instruct: —
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| LMArena Longer Query | 1366 | — |
Writing & Preference Not comparable
GPT-5.4 nano: 55.7 (#142), Qwen2.5 32B Instruct: —
| Benchmark | GPT-5.4 nano | Qwen2.5 32B Instruct |
|---|---|---|
| LMArena Text | 1372 | — |
| LMArena Creative Writing | 1314 | — |
| LMArena Multi-Turn | 1382 | — |
Frequently asked questions
Is GPT-5.4 nano better than Qwen2.5 32B Instruct?
GPT-5.4 nano is the stronger model overall, scoring 41.9 to 30.1 on the Noometry Index.
Which is cheaper, GPT-5.4 nano or Qwen2.5 32B Instruct?
GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; Qwen2.5 32B Instruct lists at $0.70 and $2.80.
Is GPT-5.4 nano or Qwen2.5 32B Instruct better for coding?
GPT-5.4 nano scores higher on coding benchmarks: 43.6 versus 38.7 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 Qwen2.5 32B Instruct share?
4 benchmarks have published results for both models. GPT-5.4 nano has 40 scored results on Noometry and Qwen2.5 32B Instruct has 7.