Model comparison
DeepSeek-V3.1 vs Qwen2-72B
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 30.0 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 19 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 8 categories and Qwen2-72B in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.1 leads 43.7 to 21.2.
- The biggest single-benchmark swing is WeirdML: 38.4% for DeepSeek-V3.1 and 11.3% for Qwen2-72B.
Side by side
| DeepSeek-V3.1 | Qwen2-72B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 42.8 | 30.0 |
| Released | 2025-08-21 | 2024-06-07 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $0.95 | — |
| Results tracked | 27 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1 leads
DeepSeek-V3.1: 40.3 (#144), Qwen2-72B: 29.1 (#310)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| WeirdML | 38.4% | 11.3% |
| LMArena Coding | 1417 | 1196 |
| BigCodeBench Instruct | — | 38.5% |
| BigCodeBench Complete | — | 54% |
Agentic & Tool Use Not comparable
DeepSeek-V3.1: —, Qwen2-72B: 17.0 (#146)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| TheAgentCompany | — | 1.1% |
| METR Time Horizons | — | 29.9% |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), Qwen2-72B: 23.2 (#181)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1417 | 1191 |
| Epoch Capabilities Index | 139.92 | 125.28 |
| SimpleBench | 40% | — |
| Kagi LLM Benchmark | 53.2% | — |
| DTBench | 82.7% | — |
| LMCA | 24.3% | — |
| ForecastBench | 58 | — |
Math DeepSeek-V3.1 leads
DeepSeek-V3.1: 38.9 (#122), Qwen2-72B: 30.2 (#236)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Math | 1420 | 1235 |
| MATH Level 5 | — | 39.1% |
Knowledge DeepSeek-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), Qwen2-72B: 21.2 (#275)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Expert | 1405 | 1171 |
| GPQA Diamond | — | 40.8% |
| Vectara Hallucination Rate | 5.5% | — |
| MMLU | — | 82.4% |
Multilingual DeepSeek-V3.1 leads
DeepSeek-V3.1: 51.6 (#106), Qwen2-72B: 35.9 (#244)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1400 | 1176 |
| LMArena Chinese | 1469 | 1240 |
| LMArena French | 1447 | 1170 |
| LMArena German | 1411 | 1151 |
| LMArena Japanese | 1378 | 1111 |
| LMArena Korean | 1337 | 1083 |
| LMArena Russian | 1405 | 1169 |
| LMArena Spanish | 1431 | 1169 |
Instruction Following DeepSeek-V3.1 leads
DeepSeek-V3.1: 73.9 (#110), Qwen2-72B: 61.7 (#241)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1400 | 1181 |
Long Context Too close to call
DeepSeek-V3.1: 36.3 (#232), Qwen2-72B: 36.1 (#235)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1422 | 1192 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference DeepSeek-V3.1 leads
DeepSeek-V3.1: 60.3 (#98), Qwen2-72B: 40.8 (#241)
| Benchmark | DeepSeek-V3.1 | Qwen2-72B |
|---|---|---|
| LMArena Text | 1420 | 1203 |
| LMArena Creative Writing | 1401 | 1181 |
| LMArena Multi-Turn | 1408 | 1196 |
| EQ-Bench Creative Writing | 1436 | — |
Frequently asked questions
Is DeepSeek-V3.1 better than Qwen2-72B?
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 30.0 on the Noometry Index.
Is DeepSeek-V3.1 or Qwen2-72B better for coding?
DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 29.1 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.1 and Qwen2-72B share?
19 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Qwen2-72B has 26.