Model comparison
Deepseek Coder v2 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 35.9 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 17 benchmarks with published results for both. Deepseek Coder v2 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 73.2 to 34.9.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.
Side by side
| Deepseek Coder v2 | Qwen3.8 Max | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 35.9 | 56.8 |
| Released | 2024-06-17 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | — | 1M |
| Max output | — | 131K |
| Input $ / M tokens | — | $2 |
| Output $ / M tokens | — | $6 |
| Results tracked | 24 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
Deepseek Coder v2: 38.1 (#183), Qwen3.8 Max: 53.5 (#29)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1251 | 1502 |
| DeepSWE | — | 57.5% |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Agentic & Tool Use Not comparable
Deepseek Coder v2: —, Qwen3.8 Max: 45.4 (#14)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
Deepseek Coder v2: 23.6 (#176), Qwen3.8 Max: 54.4 (#26)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1496 |
| NYT Connections (extended) | — | 88.3% |
| CritPt | — | 20% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LMCA | — | 46.2% |
| Epoch Capabilities Index | — | 156.41 |
| WinoGrande | 83.7% | — |
Math Qwen3.8 Max leads
Deepseek Coder v2: 34.9 (#190), Qwen3.8 Max: 73.2 (#20)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Math | 1241 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |
| GSM8K | 94.5% | — |
Knowledge Qwen3.8 Max leads
Deepseek Coder v2: 32.3 (#212), Qwen3.8 Max: 61.7 (#27)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1181 | 1507 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |
| ARC (AI2) Challenge | 64.3% | — |
Multimodal Not comparable
Deepseek Coder v2: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
Deepseek Coder v2: 36.3 (#240), Qwen3.8 Max: 56.7 (#18)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1182 | 1472 |
| LMArena Chinese | 1201 | 1538 |
| LMArena French | 1185 | 1503 |
| LMArena German | 1164 | 1483 |
| LMArena Japanese | 1126 | 1467 |
| LMArena Korean | 1104 | 1461 |
| LMArena Russian | 1188 | 1481 |
| LMArena Spanish | 1153 | 1492 |
Instruction Following Qwen3.8 Max leads
Deepseek Coder v2: 61.7 (#242), Qwen3.8 Max: 77.6 (#17)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1180 | 1479 |
Long Context Qwen3.8 Max leads
Deepseek Coder v2: 37.0 (#224), Qwen3.8 Max: 45.6 (#31)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1219 | 1489 |
Writing & Preference Qwen3.8 Max leads
Deepseek Coder v2: 38.2 (#253), Qwen3.8 Max: 67.1 (#30)
| Benchmark | Deepseek Coder v2 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1191 | 1483 |
| LMArena Creative Writing | 1120 | 1479 |
| LMArena Multi-Turn | 1177 | 1489 |
Frequently asked questions
Is Deepseek Coder v2 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 35.9 on the Noometry Index.
Is Deepseek Coder v2 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 38.1 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and Qwen3.8 Max share?
17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Qwen3.8 Max has 39.