# GPT-4.5 vs Qwen3 32B

> Qwen3 32B is the stronger model overall, scoring 39.2 to 37.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-4-5-vs-qwen3-32b
- Last updated: 2026-10-12
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. GPT-4.5 scores higher in 4 categories and Qwen3 32B in 5 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 32B leads 40.0 to 32.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for GPT-4.5 and 66.9% for Qwen3 32B.
- Qwen3 32B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-4.5 | Qwen3 32B |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 37.2 | 39.2 |
| Rank | 208 | 172 |
| Context | — | 131K |
| Input $/M | — | $0.70 |
| Output $/M | — | $2.80 |
| Weights | Proprietary | Open |

## Coding

- GPT-4.5: 42.2 (#109)
- Qwen3 32B: 37.7 (#190)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| Aider Polyglot | 44.9% | 40% |
| LMArena Coding | 1396 | 1358 |
| SciCode | — | 35.4% |
| WeirdML | 39.4% | — |
| LiveBench Coding | 75.2% | — |

## Agentic & Tool Use

- GPT-4.5: 27.9 (#97)
- Qwen3 32B: 32.6 (#62)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 48.7% |
| Cybench | 17.5% | — |

## Reasoning

- GPT-4.5: 13.9 (#330)
- Qwen3 32B: 20.2 (#241)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| LMArena Hard Prompts | 1403 | 1334 |
| Epoch Capabilities Index | 136.74 | 138.51 |
| ARC-AGI-2 | 0.8% | — |
| SimpleBench | 34.5% | — |
| Kagi LLM Benchmark | — | 54.9% |
| ARC-AGI-1 | 10.3% | — |
| CritPt | — | 0.3% |
| Chess Puzzles | — | 5% |
| EnigmaEval | 3.2% | — |
| LiveBench Reasoning | 71.1% | — |
| DTBench | — | 67.5% |
| LiveBench Data Analysis | 64.3% | — |
| LMCA | — | 17.3% |
| ForecastBench | 61.7 | — |
| LiveBench | 69% | — |

## Math

- GPT-4.5: 32.6 (#211)
- Qwen3 32B: 39.7 (#99)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 66.9% |
| LMArena Math | 1412 | 1399 |
| LiveBench Math | 69.3% | — |
| MATH Level 5 | 78.6% | — |

## Knowledge

- GPT-4.5: 32.5 (#211)
- Qwen3 32B: 40.0 (#125)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| GPQA Diamond | 68.7% | 65.7% |
| LMArena Expert | 1394 | 1362 |
| Humanity's Last Exam | 5.4% | — |
| Confabulations | 13.6% | — |
| Vectara Hallucination Rate | — | 5.9% |

## Multimodal

- GPT-4.5: 37.6 (#71)
- Qwen3 32B: —

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| LMArena Vision | 1195 | — |
| VPCT | 45% | — |

## Multilingual

- GPT-4.5: 52.5 (#83)
- Qwen3 32B: 45.6 (#167)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| LMArena Non-English | 1413 | 1317 |
| LMArena Chinese | 1421 | 1357 |
| LMArena German | 1457 | 1341 |
| LMArena Russian | 1419 | 1311 |
| LMArena French | 1418 | — |
| LMArena Japanese | 1416 | — |
| LMArena Korean | 1392 | — |

## Instruction Following

- GPT-4.5: 72.6 (#134)
- Qwen3 32B: 68.9 (#179)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| LMArena Instruction Following | 1404 | 1305 |
| LiveBench Instruction Following | 72.3% | — |

## Long Context

- GPT-4.5: 40.4 (#155)
- Qwen3 32B: 43.8 (#87)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| Fiction.LiveBench | 63.9% | 74.2% |
| LMArena Longer Query | 1406 | 1327 |

## Writing & Preference

- GPT-4.5: 56.9 (#134)
- Qwen3 32B: 52.9 (#163)

| Benchmark | GPT-4.5 | Qwen3 32B |
|---|---|---|
| LMArena Text | 1417 | 1340 |
| LMArena Creative Writing | 1394 | 1297 |
| LMArena Multi-Turn | 1444 | 1331 |
| Short-Story Creative Writing | 75.6% | — |
| EQ-Bench Creative Writing | 1258 | — |
| LiveBench Language | 61.5% | — |

## FAQ

### Is GPT-4.5 better than Qwen3 32B?

Qwen3 32B is the stronger model overall, scoring 39.2 to 37.2 on the Noometry Index.

### Is GPT-4.5 or Qwen3 32B better for coding?

GPT-4.5 scores higher on coding benchmarks: 42.2 versus 37.7 in the Noometry coding category.

### How many benchmarks do GPT-4.5 and Qwen3 32B share?

18 benchmarks have published results for both models. GPT-4.5 has 42 scored results on Noometry and Qwen3 32B has 26.