# DeepSeek V4 Pro vs Qwen3.8 27B

> DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 46.0 on the Noometry Index. Qwen3.8 27B costs 1.6× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/deepseek-v4-pro-vs-qwen3-8-27b
- Last updated: 2026-10-12
- Shared benchmarks: 30

## Summary

- They share 30 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 7 categories and Qwen3.8 27B in 2 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4 Pro leads 64.8 to 37.1.
- The biggest single-benchmark swing is NYT Connections (extended): 91.3% for DeepSeek V4 Pro and 54.5% for Qwen3.8 27B.
- Qwen3.8 27B is cheaper at $0.04 / $2.30 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 262K.

## Snapshot

| | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 54.3 | 46.0 |
| Rank | 31 | 68 |
| Context | 1M | 262K |
| Input $/M | $0.66 | $0.04 |
| Output $/M | $1.98 | $2.30 |
| Weights | Open | Open |

## Coding

- DeepSeek V4 Pro: 52.4 (#34)
- Qwen3.8 27B: 50.5 (#44)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena WebDev | 1582 | 1593 |
| SciCode | 51% | 46.6% |
| LMArena Coding | 1470 | 1482 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |

## Agentic & Tool Use

- DeepSeek V4 Pro: 32.8 (#58)
- Qwen3.8 27B: 32.9 (#57)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| APEX-Agents | 47.3% | 47.5% |
| Vending-Bench 2 | 3,285 | — |

## Reasoning

- DeepSeek V4 Pro: 56.5 (#24)
- Qwen3.8 27B: 41.0 (#54)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| ARC-AGI-2 | 61.3% | 42.4% |
| NYT Connections (extended) | 91.3% | 54.5% |
| ARC-AGI-1 | 90.5% | 87.5% |
| CritPt | 18% | 5.4% |
| LMArena Hard Prompts | 1461 | 1460 |
| DTBench | 93.9% | 88% |
| LMCA | 45.5% | 41.4% |
| Surface Evolver Bench | 40% | 45% |
| Epoch Capabilities Index | 155.31 | 149.38 |
| Kagi LLM Benchmark | 53.5% | — |
| Chess Puzzles | 47% | — |
| Mystery Game Puzzles | 43% | — |
| ForecastBench | 56.1 | — |

## Math

- DeepSeek V4 Pro: 64.8 (#30)
- Qwen3.8 27B: 37.1 (#161)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| ProofBench | 50% | 16% |
| LMArena Math | 1455 | 1456 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| OTIS Mock AIME 2024-2025 | 98.6% | — |

## Knowledge

- DeepSeek V4 Pro: 59.5 (#31)
- Qwen3.8 27B: 41.6 (#109)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena Expert | 1464 | 1482 |
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |

## Multimodal

- DeepSeek V4 Pro: —
- Qwen3.8 27B: 41.3 (#37)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena Vision | — | 1271 |

## Multilingual

- DeepSeek V4 Pro: 54.4 (#45)
- Qwen3.8 27B: 53.7 (#60)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena Non-English | 1439 | 1430 |
| LMArena Chinese | 1486 | 1504 |
| LMArena French | 1472 | 1465 |
| LMArena German | 1458 | 1438 |
| LMArena Japanese | 1445 | 1384 |
| LMArena Korean | 1447 | 1393 |
| LMArena Russian | 1453 | 1415 |
| LMArena Spanish | 1458 | 1448 |

## Instruction Following

- DeepSeek V4 Pro: 76.1 (#47)
- Qwen3.8 27B: 75.8 (#53)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena Instruction Following | 1448 | 1439 |

## Long Context

- DeepSeek V4 Pro: 45.0 (#51)
- Qwen3.8 27B: 44.3 (#70)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena Longer Query | 1458 | 1450 |
| CL-bench Life | 13.5% | — |

## Writing & Preference

- DeepSeek V4 Pro: 65.5 (#46)
- Qwen3.8 27B: 65.8 (#43)

| Benchmark | DeepSeek V4 Pro | Qwen3.8 27B |
|---|---|---|
| LMArena Text | 1451 | 1441 |
| LMArena Creative Writing | 1446 | 1384 |
| EQ-Bench Creative Writing | 1553 | 1671 |
| LMArena Multi-Turn | 1467 | 1441 |
| EQ-Bench 4 | 1166 | — |

## FAQ

### Is DeepSeek V4 Pro better than Qwen3.8 27B?

DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 46.0 on the Noometry Index. Qwen3.8 27B costs 1.6× less per token, which makes it the better buy when DeepSeek V4 Pro's lead doesn't matter for your workload.

### Which is cheaper, DeepSeek V4 Pro or Qwen3.8 27B?

Qwen3.8 27B is cheaper. It lists at $0.04 per million input tokens and $2.30 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.

### Is DeepSeek V4 Pro or Qwen3.8 27B better for coding?

DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 50.5 in the Noometry coding category.

### Which has the bigger context window?

DeepSeek V4 Pro does, with 1M tokens against 262K.

### How many benchmarks do DeepSeek V4 Pro and Qwen3.8 27B share?

30 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Qwen3.8 27B has 31.