# DeepSeek V4 Flash vs QwQ-32B

> DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 39.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-flash-vs-qwq-32b
- Last updated: 2026-10-10
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 7 categories and QwQ-32B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 23.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 94.4% for DeepSeek V4 Flash and 59.2% for QwQ-32B.

## Snapshot

| | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 53.6 | 39.8 |
| Rank | 35 | 159 |
| Context | 1M | — |
| Input $/M | $0.15 | — |
| Output $/M | $0.60 | — |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Flash: 47.9 (#59)
- QwQ-32B: 35.4 (#226)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| LMArena Coding | 1457 | 1333 |
| FrontierCode | 18.8% | — |
| Aider Polyglot | — | 20.9% |
| LMArena WebDev | 1582 | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
| BigCodeBench Instruct | — | 44.6% |
| LiveBench Coding | — | 72.2% |
| BigCodeBench Complete | — | 54.4% |
| ALE-Bench | 1,306 | — |

## Reasoning

- DeepSeek V4 Flash: 53.7 (#30)
- QwQ-32B: 23.7 (#174)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| Chess Puzzles | 33% | 5% |
| LMArena Hard Prompts | 1444 | 1325 |
| Epoch Capabilities Index | 154.49 | 137.6 |
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | 52.2% | — |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | 89% | — |
| CritPt | 16.6% | — |
| LiveBench Reasoning | — | 83.5% |
| Mystery Game Puzzles | 34% | — |
| DTBench | 90.9% | — |
| LiveBench Data Analysis | — | 65% |
| LMCA | 41.7% | — |
| ForecastBench | — | 58.3 |
| LiveBench | — | 72% |

## Math

- DeepSeek V4 Flash: 60.3 (#37)
- QwQ-32B: 38.0 (#143)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 94.4% | 59.2% |
| LMArena Math | 1427 | 1359 |
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| ProofBench | 56% | — |
| LiveBench Math | — | 77.8% |

## Knowledge

- DeepSeek V4 Flash: 55.4 (#48)
- QwQ-32B: 37.2 (#158)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| GPQA Diamond | 91% | 65.3% |
| LMArena Expert | 1441 | 1324 |
| SimpleQA Verified | 33.6% | — |
| Confabulations | — | 15.6% |

## Multilingual

- DeepSeek V4 Flash: 53.0 (#72)
- QwQ-32B: 44.8 (#176)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| LMArena Non-English | 1420 | 1305 |
| LMArena Chinese | 1468 | 1378 |
| LMArena French | 1439 | 1336 |
| LMArena German | 1418 | 1313 |
| LMArena Japanese | 1406 | 1262 |
| LMArena Korean | 1384 | 1279 |
| LMArena Russian | 1428 | 1297 |
| LMArena Spanish | 1436 | 1354 |

## Instruction Following

- DeepSeek V4 Flash: 74.9 (#81)
- QwQ-32B: 72.6 (#137)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| LMArena Instruction Following | 1421 | 1297 |
| LiveBench Instruction Following | — | 81.8% |

## Long Context

- DeepSeek V4 Flash: 43.8 (#85)
- QwQ-32B: 49.0 (#11)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| LMArena Longer Query | 1434 | 1308 |
| Fiction.LiveBench | — | 83.3% |

## Writing & Preference

- DeepSeek V4 Flash: 63.8 (#61)
- QwQ-32B: 50.6 (#180)

| Benchmark | DeepSeek V4 Flash | QwQ-32B |
|---|---|---|
| LMArena Text | 1432 | 1329 |
| LMArena Creative Writing | 1403 | 1288 |
| EQ-Bench Creative Writing | 1559 | 1257 |
| LMArena Multi-Turn | 1449 | 1314 |
| Short-Story Creative Writing | — | 80.2% |
| LiveBench Language | — | 51.4% |

## FAQ

### Is DeepSeek V4 Flash better than QwQ-32B?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 39.8 on the Noometry Index.

### Is DeepSeek V4 Flash or QwQ-32B better for coding?

DeepSeek V4 Flash scores higher on coding benchmarks: 47.9 versus 35.4 in the Noometry coding category.

### How many benchmarks do DeepSeek V4 Flash and QwQ-32B share?

22 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and QwQ-32B has 36.