# DeepSeek V4 Flash vs Qwen3-Next 80B-A3B Instruct

> DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 43.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-flash-vs-qwen3-next-80b-a3b-instruct
- Last updated: 2026-10-12
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. DeepSeek V4 Flash scores higher in 8 categories and Qwen3-Next 80B-A3B Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 31.1.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 52.2% for DeepSeek V4 Flash and 66.7% for Qwen3-Next 80B-A3B Instruct.
- DeepSeek V4 Flash is cheaper at $0.15 / $0.60 per million input/output tokens, against $0.50 / $2 for Qwen3-Next 80B-A3B Instruct.
- DeepSeek V4 Flash accepts more context: 1M tokens versus 131K.

## Snapshot

| | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 53.6 | 43.0 |
| Rank | 35 | 102 |
| Context | 1M | 131K |
| Input $/M | $0.15 | $0.50 |
| Output $/M | $0.60 | $2 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Flash: 47.9 (#59)
- Qwen3-Next 80B-A3B Instruct: 42.5 (#98)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Coding | 1457 | 1440 |
| FrontierCode | 18.8% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 49.9% | — |
| WeirdML | 63% | — |
| ALE-Bench | 1,306 | — |

## Reasoning

- DeepSeek V4 Flash: 53.7 (#30)
- Qwen3-Next 80B-A3B Instruct: 31.1 (#81)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 52.2% | 66.7% |
| LMArena Hard Prompts | 1444 | 1428 |
| ARC-AGI-2 | 61.4% | — |
| SimpleBench | 61.1% | — |
| NYT Connections (extended) | 89.6% | — |
| ARC-AGI-1 | 89% | — |
| CritPt | 16.6% | — |
| Chess Puzzles | 33% | — |
| Mystery Game Puzzles | 34% | — |
| DTBench | 90.9% | — |
| LMCA | 41.7% | — |
| Epoch Capabilities Index | 154.49 | — |

## Math

- DeepSeek V4 Flash: 60.3 (#37)
- Qwen3-Next 80B-A3B Instruct: 38.8 (#126)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Math | 1427 | 1440 |
| FrontierMath (Tiers 1-3) | 57.5% | — |
| FrontierMath Tier 4 | 24.4% | — |
| MathArena Final-Answer Competitions | 76.5% | — |
| OTIS Mock AIME 2024-2025 | 94.4% | — |
| ProofBench | 56% | — |
| Omni-MATH | — | 46.7% |

## Knowledge

- DeepSeek V4 Flash: 55.4 (#48)
- Qwen3-Next 80B-A3B Instruct: 41.8 (#106)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Expert | 1441 | 1417 |
| GPQA Diamond | 91% | — |
| SimpleQA Verified | 33.6% | — |
| MMLU-Pro | — | 78.6% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 63% |

## Multilingual

- DeepSeek V4 Flash: 53.0 (#72)
- Qwen3-Next 80B-A3B Instruct: 52.1 (#93)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Non-English | 1420 | 1407 |
| LMArena Chinese | 1468 | 1460 |
| LMArena French | 1439 | 1413 |
| LMArena German | 1418 | 1417 |
| LMArena Japanese | 1406 | 1395 |
| LMArena Korean | 1384 | 1364 |
| LMArena Russian | 1428 | 1404 |
| LMArena Spanish | 1436 | 1435 |

## Instruction Following

- DeepSeek V4 Flash: 74.9 (#81)
- Qwen3-Next 80B-A3B Instruct: 70.8 (#159)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Instruction Following | 1421 | 1389 |
| IFEval | — | 81% |

## Long Context

- DeepSeek V4 Flash: 43.8 (#85)
- Qwen3-Next 80B-A3B Instruct: 37.0 (#223)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Longer Query | 1434 | 1403 |
| Fiction.LiveBench | — | 55.6% |

## Writing & Preference

- DeepSeek V4 Flash: 63.8 (#61)
- Qwen3-Next 80B-A3B Instruct: 58.0 (#121)

| Benchmark | DeepSeek V4 Flash | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Text | 1432 | 1417 |
| LMArena Creative Writing | 1403 | 1334 |
| LMArena Multi-Turn | 1449 | 1416 |
| EQ-Bench Creative Writing | 1559 | — |
| WildBench | — | 80.7% |

## FAQ

### Is DeepSeek V4 Flash better than Qwen3-Next 80B-A3B Instruct?

DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 43.0 on the Noometry Index.

### Which is cheaper, DeepSeek V4 Flash or Qwen3-Next 80B-A3B Instruct?

DeepSeek V4 Flash is cheaper. It lists at $0.15 per million input tokens and $0.60 per million output tokens; Qwen3-Next 80B-A3B Instruct lists at $0.50 and $2.

### Is DeepSeek V4 Flash or Qwen3-Next 80B-A3B Instruct better for coding?

DeepSeek V4 Flash scores higher on coding benchmarks: 47.9 versus 42.5 in the Noometry coding category.

### Which has the bigger context window?

DeepSeek V4 Flash does, with 1M tokens against 131K.

### How many benchmarks do DeepSeek V4 Flash and Qwen3-Next 80B-A3B Instruct share?

18 benchmarks have published results for both models. DeepSeek V4 Flash has 41 scored results on Noometry and Qwen3-Next 80B-A3B Instruct has 25.