Model comparison
DeepSeek-R1 vs Qwen2-72B
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 30.0 on the Noometry Index.
Last verified . 22 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. DeepSeek-R1 scores higher in 8 categories and Qwen2-72B in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-R1 leads 44.5 to 21.2.
- The biggest single-benchmark swing is MATH Level 5: 96.6% for DeepSeek-R1 and 39.1% for Qwen2-72B.
- Qwen2-72B has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-R1 | Qwen2-72B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 42.3 | 30.0 |
| Released | 2025-01-20 | 2024-06-07 |
| Weights | Proprietary | Open |
| Context window | 164K | — |
| Max output | 64K | — |
| Input $ / M tokens | $0.50 | — |
| Output $ / M tokens | $2.15 | — |
| Results tracked | 52 | 26 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1 leads
DeepSeek-R1: 46.3 (#68), Qwen2-72B: 29.1 (#310)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| WeirdML | 41.6% | 11.3% |
| LMArena Coding | 1427 | 1196 |
| Aider Polyglot | 71.4% | — |
| SciCode | 35.7% | — |
| BigCodeBench Instruct | — | 38.5% |
| LiveBench Coding | 66.7% | — |
| BigCodeBench Complete | — | 54% |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use DeepSeek-R1 leads
DeepSeek-R1: 30.7 (#75), Qwen2-72B: 17.0 (#146)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| METR Time Horizons | 53.8% | 29.9% |
| TheAgentCompany | — | 1.1% |
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
Reasoning Qwen2-72B leads
DeepSeek-R1: 18.6 (#278), Qwen2-72B: 23.2 (#181)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| LMArena Hard Prompts | 1416 | 1191 |
| Epoch Capabilities Index | 141.29 | 125.28 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 40.8% | — |
| Kagi LLM Benchmark | 69.4% | — |
| ARC-AGI-1 | 21.2% | — |
| CritPt | 1.1% | — |
| LiveBench Reasoning | 83.2% | — |
| LiveBench Data Analysis | 69.8% | — |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |
Math DeepSeek-R1 leads
DeepSeek-R1: 43.8 (#79), Qwen2-72B: 30.2 (#236)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| LMArena Math | 1400 | 1235 |
| MATH Level 5 | 96.6% | 39.1% |
| OTIS Mock AIME 2024-2025 | 66.4% | — |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
Knowledge DeepSeek-R1 leads
DeepSeek-R1: 44.5 (#87), Qwen2-72B: 21.2 (#275)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| GPQA Diamond | 76.3% | 40.8% |
| LMArena Expert | 1394 | 1171 |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
| MMLU | — | 82.4% |
Multilingual DeepSeek-R1 leads
DeepSeek-R1: 52.4 (#85), Qwen2-72B: 35.9 (#244)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| LMArena Non-English | 1412 | 1176 |
| LMArena Chinese | 1442 | 1240 |
| LMArena French | 1417 | 1170 |
| LMArena German | 1404 | 1151 |
| LMArena Japanese | 1391 | 1111 |
| LMArena Korean | 1360 | 1083 |
| LMArena Russian | 1423 | 1169 |
| LMArena Spanish | 1411 | 1169 |
Instruction Following DeepSeek-R1 leads
DeepSeek-R1: 72.0 (#143), Qwen2-72B: 61.7 (#241)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| LMArena Instruction Following | 1382 | 1181 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |
Long Context DeepSeek-R1 leads
DeepSeek-R1: 45.4 (#36), Qwen2-72B: 36.1 (#235)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| LMArena Longer Query | 1391 | 1192 |
| Fiction.LiveBench | 75% | — |
Writing & Preference DeepSeek-R1 leads
DeepSeek-R1: 61.4 (#88), Qwen2-72B: 40.8 (#241)
| Benchmark | DeepSeek-R1 | Qwen2-72B |
|---|---|---|
| LMArena Text | 1428 | 1203 |
| LMArena Creative Writing | 1405 | 1181 |
| LMArena Multi-Turn | 1405 | 1196 |
| Short-Story Creative Writing | 83% | — |
| EQ-Bench Creative Writing | 1500 | — |
| WildBench | 82.8% | — |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than Qwen2-72B?
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 30.0 on the Noometry Index.
Is DeepSeek-R1 or Qwen2-72B better for coding?
DeepSeek-R1 scores higher on coding benchmarks: 46.3 versus 29.1 in the Noometry coding category.
How many benchmarks do DeepSeek-R1 and Qwen2-72B share?
22 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Qwen2-72B has 26.