# Qwen3.8 Max vs QwQ-32B

> Qwen3.8 Max is the stronger model overall, scoring 56.0 to 39.4 on the Noometry Index.

- Canonical page: https://noometry.com/compare/qwen3-8-max-vs-qwq-32b
- Last updated: 2026-10-10
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Qwen3.8 Max scores higher in 7 categories and QwQ-32B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.1 to 36.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for Qwen3.8 Max and 59.2% for QwQ-32B.
- QwQ-32B has downloadable open weights; the other is API-only.

## Snapshot

| | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 56.0 | 39.4 |
| Rank | 21 | 153 |
| Context | 1M | — |
| Input $/M | $2 | — |
| Output $/M | $6 | — |
| Weights | Proprietary | Open |

## Coding

- Qwen3.8 Max: 52.3 (#30)
- QwQ-32B: 34.9 (#220)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| LMArena Coding | 1499 | 1333 |
| DeepSWE | 57.5% | — |
| Aider Polyglot | — | 20.9% |
| LMArena WebDev | 1671 | — |
| FrontierSWE | 15.8% | — |
| SciCode | 52.9% | — |
| BigCodeBench Instruct | — | 44.6% |
| LiveBench Coding | — | 72.2% |
| BigCodeBench Complete | — | 54.4% |

## Agentic & Tool Use

- Qwen3.8 Max: 42.6 (#17)
- QwQ-32B: —

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| τ²-bench Banking | 55.1% | — |
| GDP.pdf | 23.2% | — |

## Reasoning

- Qwen3.8 Max: 53.4 (#26)
- QwQ-32B: 22.9 (#171)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| Chess Puzzles | 40% | 5% |
| LMArena Hard Prompts | 1493 | 1326 |
| Epoch Capabilities Index | 156.69 | 137.6 |
| NYT Connections (extended) | 88.3% | — |
| CritPt | 20% | — |
| LiveBench Reasoning | — | 83.5% |
| Mystery Game Puzzles | 38% | — |
| DTBench | 92% | — |
| LiveBench Data Analysis | — | 65% |
| LMCA | 46.2% | — |
| ForecastBench | — | 58.3 |
| LiveBench | — | 72% |

## Math

- Qwen3.8 Max: 73.1 (#15)
- QwQ-32B: 36.2 (#123)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 100% | 59.2% |
| LMArena Math | 1498 | 1359 |
| FrontierMath (Tiers 1-3) | 74.7% | — |
| FrontierMath Tier 4 | 46.3% | — |
| ProofBench | 58% | — |
| LiveBench Math | — | 77.8% |

## Knowledge

- Qwen3.8 Max: 61.7 (#25)
- QwQ-32B: 37.2 (#150)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| GPQA Diamond | 92.7% | 65.3% |
| LMArena Expert | 1504 | 1323 |
| SimpleQA Verified | 47.3% | — |
| Confabulations | — | 15.6% |

## Multimodal

- Qwen3.8 Max: 36.9 (#70)
- QwQ-32B: —

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| LMArena Vision | 1314 | — |
| Furniture Assembly | 20% | — |

## Multilingual

- Qwen3.8 Max: 56.6 (#18)
- QwQ-32B: 45.0 (#171)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| LMArena Non-English | 1468 | 1306 |
| LMArena Chinese | 1537 | 1378 |
| LMArena French | 1502 | 1337 |
| LMArena German | 1489 | 1313 |
| LMArena Japanese | 1466 | 1263 |
| LMArena Korean | 1467 | 1278 |
| LMArena Russian | 1478 | 1297 |
| LMArena Spanish | 1495 | 1353 |

## Instruction Following

- Qwen3.8 Max: 77.7 (#19)
- QwQ-32B: 72.8 (#131)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| LMArena Instruction Following | 1478 | 1297 |
| LiveBench Instruction Following | — | 81.8% |

## Long Context

- Qwen3.8 Max: 45.5 (#30)
- QwQ-32B: 49.0 (#11)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| LMArena Longer Query | 1486 | 1308 |
| Fiction.LiveBench | — | 83.3% |

## Writing & Preference

- Qwen3.8 Max: 67.0 (#31)
- QwQ-32B: 50.9 (#175)

| Benchmark | Qwen3.8 Max | QwQ-32B |
|---|---|---|
| LMArena Text | 1480 | 1329 |
| LMArena Creative Writing | 1478 | 1288 |
| LMArena Multi-Turn | 1487 | 1314 |
| Short-Story Creative Writing | — | 80.2% |
| EQ-Bench Creative Writing | — | 1257 |
| LiveBench Language | — | 51.4% |

## FAQ

### Is Qwen3.8 Max better than QwQ-32B?

Qwen3.8 Max is the stronger model overall, scoring 56.0 to 39.4 on the Noometry Index.

### Is Qwen3.8 Max or QwQ-32B better for coding?

Qwen3.8 Max scores higher on coding benchmarks: 52.3 versus 34.9 in the Noometry coding category.

### How many benchmarks do Qwen3.8 Max and QwQ-32B share?

21 benchmarks have published results for both models. Qwen3.8 Max has 38 scored results on Noometry and QwQ-32B has 36.