Model comparison
DeepSeek-V3.1 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 42.8 on the Noometry Index. DeepSeek-V3.1 costs 7.1× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. DeepSeek-V3.1 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 38.9.
- The biggest single-benchmark swing is LMCA: 24.3% for DeepSeek-V3.1 and 46.2% for Qwen3.8 Max.
- DeepSeek-V3.1 is cheaper at $0.25 / $0.95 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Qwen3.8 Max accepts more context: 1M tokens versus 164K.
- DeepSeek-V3.1 has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-V3.1 | Qwen3.8 Max | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 42.8 | 56.8 |
| Released | 2025-08-21 | 2026-08-02 |
| Weights | Open | Proprietary |
| Context window | 164K | 1M |
| Max output | 8K | 131K |
| Input $ / M tokens | $0.25 | $2 |
| Output $ / M tokens | $0.95 | $6 |
| Results tracked | 27 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
DeepSeek-V3.1: 40.3 (#144), Qwen3.8 Max: 53.5 (#29)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Coding | 1417 | 1502 |
| DeepSWE | — | 57.5% |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| SciCode | — | 53.2% |
| WeirdML | 38.4% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3.1: —, Qwen3.8 Max: 45.4 (#14)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| GDP.pdf | — | 23.2% |
Reasoning Qwen3.8 Max leads
DeepSeek-V3.1: 27.9 (#110), Qwen3.8 Max: 54.4 (#26)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Hard Prompts | 1417 | 1496 |
| DTBench | 82.7% | 92% |
| LMCA | 24.3% | 46.2% |
| Epoch Capabilities Index | 139.92 | 156.41 |
| SimpleBench | 40% | — |
| Kagi LLM Benchmark | 53.2% | — |
| NYT Connections (extended) | — | 88.3% |
| CritPt | — | 20% |
| Chess Puzzles | — | 40% |
| Mystery Game Puzzles | — | 38% |
| ForecastBench | 58 | — |
Math Qwen3.8 Max leads
DeepSeek-V3.1: 38.9 (#122), Qwen3.8 Max: 73.2 (#20)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Math | 1420 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| OTIS Mock AIME 2024-2025 | — | 100% |
| ProofBench | — | 58% |
Knowledge Qwen3.8 Max leads
DeepSeek-V3.1: 43.7 (#90), Qwen3.8 Max: 61.7 (#27)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Expert | 1405 | 1507 |
| GPQA Diamond | — | 92.7% |
| SimpleQA Verified | — | 47.3% |
| Vectara Hallucination Rate | 5.5% | — |
Multimodal Not comparable
DeepSeek-V3.1: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
DeepSeek-V3.1: 51.6 (#106), Qwen3.8 Max: 56.7 (#18)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1400 | 1472 |
| LMArena Chinese | 1469 | 1538 |
| LMArena French | 1447 | 1503 |
| LMArena German | 1411 | 1483 |
| LMArena Japanese | 1378 | 1467 |
| LMArena Korean | 1337 | 1461 |
| LMArena Russian | 1405 | 1481 |
| LMArena Spanish | 1431 | 1492 |
Instruction Following Qwen3.8 Max leads
DeepSeek-V3.1: 73.9 (#110), Qwen3.8 Max: 77.6 (#17)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1400 | 1479 |
Long Context Qwen3.8 Max leads
DeepSeek-V3.1: 36.3 (#232), Qwen3.8 Max: 45.6 (#31)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1422 | 1489 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference Qwen3.8 Max leads
DeepSeek-V3.1: 60.3 (#98), Qwen3.8 Max: 67.1 (#30)
| Benchmark | DeepSeek-V3.1 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1420 | 1483 |
| LMArena Creative Writing | 1401 | 1479 |
| LMArena Multi-Turn | 1408 | 1489 |
| EQ-Bench Creative Writing | 1436 | — |
Frequently asked questions
Is DeepSeek-V3.1 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 42.8 on the Noometry Index. DeepSeek-V3.1 costs 7.1× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3.1 or Qwen3.8 Max?
DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is DeepSeek-V3.1 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 40.3 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 164K.
How many benchmarks do DeepSeek-V3.1 and Qwen3.8 Max share?
20 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Qwen3.8 Max has 39.