# GPT-5.4 mini vs QwQ-32B

> GPT-5.4 mini is the stronger model overall, scoring 44.8 to 39.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-5-4-mini-vs-qwq-32b
- Last updated: 2026-10-10
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. GPT-5.4 mini scores higher in 7 categories and QwQ-32B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GPT-5.4 mini leads 51.4 to 37.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 88.9% for GPT-5.4 mini and 59.2% for QwQ-32B.
- QwQ-32B has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| Provider | OpenAI | Alibaba (Qwen) |
| Noometry Index | 44.8 | 39.4 |
| Rank | 69 | 153 |
| Context | 400K | — |
| Input $/M | $0.75 | — |
| Output $/M | $4.50 | — |
| Weights | Proprietary | Open |

## Coding

- GPT-5.4 mini: 45.0 (#67)
- QwQ-32B: 34.9 (#220)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| LMArena Coding | 1438 | 1333 |
| FrontierCode | 27% | — |
| Aider Polyglot | — | 20.9% |
| LMArena WebDev | 1397 | — |
| SciCode | 49.9% | — |
| WeirdML | 60.3% | — |
| BigCodeBench Instruct | — | 44.6% |
| LiveBench Coding | — | 72.2% |
| BigCodeBench Complete | — | 54.4% |
| ALE-Bench | 1,189 | — |

## Agentic & Tool Use

- GPT-5.4 mini: 29.6 (#74)
- QwQ-32B: —

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| DeepResearch Bench | 36.3% | — |

## Reasoning

- GPT-5.4 mini: 30.3 (#77)
- QwQ-32B: 22.9 (#171)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| Chess Puzzles | 24% | 5% |
| LMArena Hard Prompts | 1424 | 1326 |
| Epoch Capabilities Index | 148.83 | 137.6 |
| ForecastBench | 57 | 58.3 |
| ARC-AGI-2 | 18.9% | — |
| Kagi LLM Benchmark | 37.9% | — |
| NYT Connections (extended) | 61.8% | — |
| ARC-AGI-1 | 63.7% | — |
| CritPt | 10% | — |
| Thematic Generalization | 61.7% | — |
| LiveBench Reasoning | — | 83.5% |
| Mystery Game Puzzles | 11% | — |
| DTBench | 80% | — |
| LiveBench Data Analysis | — | 65% |
| LMCA | 40.8% | — |
| LiveBench | — | 72% |

## Math

- GPT-5.4 mini: 45.4 (#63)
- QwQ-32B: 36.2 (#123)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 88.9% | 59.2% |
| LMArena Math | 1418 | 1359 |
| FrontierMath (Tiers 1-3) | 51.2% | — |
| FrontierMath Tier 4 | 9.8% | — |
| ProofBench | 21% | — |
| LiveBench Math | — | 77.8% |
| FrontierMath (Feb 2025 set) | 28.3% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |

## Knowledge

- GPT-5.4 mini: 51.4 (#62)
- QwQ-32B: 37.2 (#150)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| GPQA Diamond | 86.9% | 65.3% |
| LMArena Expert | 1434 | 1323 |
| SimpleQA Verified | 29.4% | — |
| Confabulations | — | 15.6% |
| Vectara Hallucination Rate | 5.5% | — |

## Multimodal

- GPT-5.4 mini: 39.0 (#51)
- QwQ-32B: —

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| LMArena Vision | 1245 | — |

## Multilingual

- GPT-5.4 mini: 52.1 (#94)
- QwQ-32B: 45.0 (#171)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| LMArena Non-English | 1405 | 1306 |
| LMArena Chinese | 1446 | 1378 |
| LMArena French | 1440 | 1337 |
| LMArena German | 1407 | 1313 |
| LMArena Japanese | 1375 | 1263 |
| LMArena Korean | 1369 | 1278 |
| LMArena Russian | 1417 | 1297 |
| LMArena Spanish | 1403 | 1353 |

## Instruction Following

- GPT-5.4 mini: 74.3 (#99)
- QwQ-32B: 72.8 (#131)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| LMArena Instruction Following | 1405 | 1297 |
| LiveBench Instruction Following | — | 81.8% |

## Long Context

- GPT-5.4 mini: 42.9 (#107)
- QwQ-32B: 49.0 (#11)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| LMArena Longer Query | 1407 | 1308 |
| Fiction.LiveBench | — | 83.3% |

## Writing & Preference

- GPT-5.4 mini: 64.2 (#56)
- QwQ-32B: 50.9 (#175)

| Benchmark | GPT-5.4 mini | QwQ-32B |
|---|---|---|
| LMArena Text | 1411 | 1329 |
| LMArena Creative Writing | 1370 | 1288 |
| EQ-Bench Creative Writing | 1665 | 1257 |
| LMArena Multi-Turn | 1429 | 1314 |
| Short-Story Creative Writing | — | 80.2% |
| LiveBench Language | — | 51.4% |

## FAQ

### Is GPT-5.4 mini better than QwQ-32B?

GPT-5.4 mini is the stronger model overall, scoring 44.8 to 39.4 on the Noometry Index.

### Is GPT-5.4 mini or QwQ-32B better for coding?

GPT-5.4 mini scores higher on coding benchmarks: 45.0 versus 34.9 in the Noometry coding category.

### How many benchmarks do GPT-5.4 mini and QwQ-32B share?

23 benchmarks have published results for both models. GPT-5.4 mini has 46 scored results on Noometry and QwQ-32B has 36.