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

> DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 43.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v4-pro-vs-qwen3-next-80b-a3b-instruct
- Last updated: 2026-10-12
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 8 categories and Qwen3-Next 80B-A3B Instruct in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Pro leads 64.8 to 38.8.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 53.5% for DeepSeek V4 Pro and 66.7% for Qwen3-Next 80B-A3B Instruct.
- Qwen3-Next 80B-A3B Instruct is cheaper at $0.50 / $2 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 131K.

## Snapshot

| | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 54.3 | 43.0 |
| Rank | 31 | 102 |
| Context | 1M | 131K |
| Input $/M | $0.66 | $0.50 |
| Output $/M | $1.98 | $2 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Pro: 52.4 (#34)
- Qwen3-Next 80B-A3B Instruct: 42.5 (#98)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Coding | 1470 | 1440 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |

## Agentic & Tool Use

- DeepSeek V4 Pro: 32.8 (#58)
- Qwen3-Next 80B-A3B Instruct: —

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |

## Reasoning

- DeepSeek V4 Pro: 56.5 (#24)
- Qwen3-Next 80B-A3B Instruct: 31.1 (#81)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 53.5% | 66.7% |
| LMArena Hard Prompts | 1461 | 1428 |
| ARC-AGI-2 | 61.3% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| Chess Puzzles | 47% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.31 | — |
| ForecastBench | 56.1 | — |

## Math

- DeepSeek V4 Pro: 64.8 (#30)
- Qwen3-Next 80B-A3B Instruct: 38.8 (#126)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Math | 1455 | 1440 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| OTIS Mock AIME 2024-2025 | 98.6% | — |
| ProofBench | 50% | — |
| Omni-MATH | — | 46.7% |

## Knowledge

- DeepSeek V4 Pro: 59.5 (#31)
- Qwen3-Next 80B-A3B Instruct: 41.8 (#106)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| Vectara Hallucination Rate | 8.6% | 9.3% |
| LMArena Expert | 1464 | 1417 |
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
| MMLU-Pro | — | 78.6% |
| GPQA (HELM) | — | 63% |

## Multilingual

- DeepSeek V4 Pro: 54.4 (#45)
- Qwen3-Next 80B-A3B Instruct: 52.1 (#93)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Non-English | 1439 | 1407 |
| LMArena Chinese | 1486 | 1460 |
| LMArena French | 1472 | 1413 |
| LMArena German | 1458 | 1417 |
| LMArena Japanese | 1445 | 1395 |
| LMArena Korean | 1447 | 1364 |
| LMArena Russian | 1453 | 1404 |
| LMArena Spanish | 1458 | 1435 |

## Instruction Following

- DeepSeek V4 Pro: 76.1 (#47)
- Qwen3-Next 80B-A3B Instruct: 70.8 (#159)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Instruction Following | 1448 | 1389 |
| IFEval | — | 81% |

## Long Context

- DeepSeek V4 Pro: 45.0 (#51)
- Qwen3-Next 80B-A3B Instruct: 37.0 (#223)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Longer Query | 1458 | 1403 |
| Fiction.LiveBench | — | 55.6% |
| CL-bench Life | 13.5% | — |

## Writing & Preference

- DeepSeek V4 Pro: 65.5 (#46)
- Qwen3-Next 80B-A3B Instruct: 58.0 (#121)

| Benchmark | DeepSeek V4 Pro | Qwen3-Next 80B-A3B Instruct |
|---|---|---|
| LMArena Text | 1451 | 1417 |
| LMArena Creative Writing | 1446 | 1334 |
| LMArena Multi-Turn | 1467 | 1416 |
| EQ-Bench Creative Writing | 1553 | — |
| WildBench | — | 80.7% |
| EQ-Bench 4 | 1166 | — |

## FAQ

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

DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 43.0 on the Noometry Index.

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

Qwen3-Next 80B-A3B Instruct is cheaper. It lists at $0.50 per million input tokens and $2 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.

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

DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 42.5 in the Noometry coding category.

### Which has the bigger context window?

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

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

19 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Qwen3-Next 80B-A3B Instruct has 25.