# DeepSeek-V2.5 (Sep 2024) vs Qwen3.8 Max

> Qwen3.8 Max is the stronger model overall, scoring 56.0 to 37.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-v2-5-vs-qwen3-8-max
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 0 categories and Qwen3.8 Max in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3.8 Max leads 73.1 to 33.3.
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.

## Snapshot

| | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 37.0 | 56.0 |
| Rank | 197 | 21 |
| Context | — | 1M |
| Input $/M | — | $2 |
| Output $/M | — | $6 |
| Weights | Open | Proprietary |

## Coding

- DeepSeek-V2.5 (Sep 2024): 31.3 (#274)
- Qwen3.8 Max: 52.3 (#30)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1310 | 1499 |
| DeepSWE | — | 57.5% |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1671 |
| FrontierSWE | — | 15.8% |
| SciCode | — | 52.9% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |

## Agentic & Tool Use

- DeepSeek-V2.5 (Sep 2024): —
- Qwen3.8 Max: 42.6 (#17)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |

## Reasoning

- DeepSeek-V2.5 (Sep 2024): 24.5 (#145)
- Qwen3.8 Max: 53.4 (#26)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1493 |
| NYT Connections (extended) | — | 88.3% |
| CritPt | — | 20% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LMCA | — | 46.2% |
| Epoch Capabilities Index | — | 156.69 |

## Math

- DeepSeek-V2.5 (Sep 2024): 33.3 (#177)
- Qwen3.8 Max: 73.1 (#15)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Math | 1289 | 1498 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |

## Knowledge

- DeepSeek-V2.5 (Sep 2024): 34.2 (#190)
- Qwen3.8 Max: 61.7 (#25)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1266 | 1504 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |

## Multimodal

- DeepSeek-V2.5 (Sep 2024): —
- Qwen3.8 Max: 36.9 (#70)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |

## Multilingual

- DeepSeek-V2.5 (Sep 2024): 42.7 (#188)
- Qwen3.8 Max: 56.6 (#18)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1273 | 1468 |
| LMArena Chinese | 1317 | 1537 |
| LMArena French | 1288 | 1502 |
| LMArena German | 1258 | 1489 |
| LMArena Japanese | 1228 | 1466 |
| LMArena Korean | 1209 | 1467 |
| LMArena Russian | 1289 | 1478 |
| LMArena Spanish | 1248 | 1495 |

## Instruction Following

- DeepSeek-V2.5 (Sep 2024): 67.7 (#189)
- Qwen3.8 Max: 77.7 (#19)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1280 | 1478 |

## Long Context

- DeepSeek-V2.5 (Sep 2024): 39.5 (#169)
- Qwen3.8 Max: 45.5 (#30)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1301 | 1486 |

## Writing & Preference

- DeepSeek-V2.5 (Sep 2024): 50.0 (#182)
- Qwen3.8 Max: 67.0 (#31)

| Benchmark | DeepSeek-V2.5 (Sep 2024) | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1294 | 1480 |
| LMArena Creative Writing | 1285 | 1478 |
| LMArena Multi-Turn | 1297 | 1487 |

## FAQ

### Is DeepSeek-V2.5 (Sep 2024) better than Qwen3.8 Max?

Qwen3.8 Max is the stronger model overall, scoring 56.0 to 37.0 on the Noometry Index.

### Is DeepSeek-V2.5 (Sep 2024) or Qwen3.8 Max better for coding?

Qwen3.8 Max scores higher on coding benchmarks: 52.3 versus 31.3 in the Noometry coding category.

### How many benchmarks do DeepSeek-V2.5 (Sep 2024) and Qwen3.8 Max share?

17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and Qwen3.8 Max has 38.