Model comparison
DeepSeek-R1 vs Qwen2.5-VL 72B Instruct
DeepSeek-R1 is the stronger model overall, scoring 42.0 to 29.4 on the Noometry Index.
Last verified . 1 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 1 benchmark with published results for both. DeepSeek-R1 scores higher in 1 category and Qwen2.5-VL 72B Instruct in 1 category; 2 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where DeepSeek-R1 leads 30.4 to 18.4.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 69.4% for DeepSeek-R1 and 36% for Qwen2.5-VL 72B Instruct.
- DeepSeek-R1 is cheaper at $0.50 / $2.15 per million input/output tokens, against $2.80 / $8.40 for Qwen2.5-VL 72B Instruct.
- DeepSeek-R1 accepts more context: 164K tokens versus 131K.
- Qwen2.5-VL 72B Instruct has downloadable open weights; the other is API-only.
Side by side
| DeepSeek-R1 | Qwen2.5-VL 72B Instruct | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 42.0 | 29.4 |
| Released | 2025-01-20 | 2024-09 |
| Weights | Proprietary | Open |
| Context window | 164K | 131K |
| Max output | 64K | 8K |
| Input $ / M tokens | $0.50 | $2.80 |
| Output $ / M tokens | $2.15 | $8.40 |
| Results tracked | 52 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
DeepSeek-R1: 45.9 (#63), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Aider Polyglot | 71.4% | — |
| SciCode | 35.7% | — |
| WeirdML | 41.6% | — |
| LiveBench Coding | 66.7% | — |
| LMArena Coding | 1427 | — |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use DeepSeek-R1 leads
DeepSeek-R1: 30.4 (#68), Qwen2.5-VL 72B Instruct: 18.4 (#136)
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| DeepResearch Bench | 35.1% | — |
| OSWorld | — | 5% |
| BALROG | 34.9% | — |
| METR Time Horizons | 53.8% | — |
Reasoning Qwen2.5-VL 72B Instruct leads
DeepSeek-R1: 18.7 (#253), Qwen2.5-VL 72B Instruct: 20.0 (#228)
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Kagi LLM Benchmark | 69.4% | 36% |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 40.8% | — |
| ARC-AGI-1 | 21.2% | — |
| CritPt | 1.1% | — |
| LiveBench Reasoning | 83.2% | — |
| LMArena Hard Prompts | 1415 | — |
| LiveBench Data Analysis | 69.8% | — |
| Epoch Capabilities Index | 141.3 | — |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |
Math Not comparable
DeepSeek-R1: 42.0 (#70), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | — |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| LMArena Math | 1401 | — |
| MATH Level 5 | 96.6% | — |
Knowledge Not comparable
DeepSeek-R1: 44.3 (#81), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| GPQA Diamond | 76.3% | — |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
| LMArena Expert | 1394 | — |
Multimodal Not comparable
DeepSeek-R1: —, Qwen2.5-VL 72B Instruct: 32.9 (#92)
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Vision | — | 1107 |
| Video-MME | — | 73.5% |
| GeoBench | — | 62% |
| SpatialViz-Bench | — | 33.3% |
Multilingual Not comparable
DeepSeek-R1: 52.6 (#80), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Non-English | 1412 | — |
| LMArena Chinese | 1443 | — |
| LMArena French | 1415 | — |
| LMArena German | 1405 | — |
| LMArena Japanese | 1394 | — |
| LMArena Korean | 1359 | — |
| LMArena Russian | 1423 | — |
| LMArena Spanish | 1410 | — |
Instruction Following Not comparable
DeepSeek-R1: 72.2 (#138), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |
| LMArena Instruction Following | 1382 | — |
Long Context Not comparable
DeepSeek-R1: 45.4 (#36), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| Fiction.LiveBench | 75% | — |
| LMArena Longer Query | 1391 | — |
Writing & Preference Not comparable
DeepSeek-R1: 61.5 (#84), Qwen2.5-VL 72B Instruct: —
| Benchmark | DeepSeek-R1 | Qwen2.5-VL 72B Instruct |
|---|---|---|
| LMArena Text | 1428 | — |
| LMArena Creative Writing | 1406 | — |
| Short-Story Creative Writing | 83% | — |
| EQ-Bench Creative Writing | 1500 | — |
| WildBench | 82.8% | — |
| LMArena Multi-Turn | 1404 | — |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than Qwen2.5-VL 72B Instruct?
DeepSeek-R1 is the stronger model overall, scoring 42.0 to 29.4 on the Noometry Index.
Which is cheaper, DeepSeek-R1 or Qwen2.5-VL 72B Instruct?
DeepSeek-R1 is cheaper. It lists at $0.50 per million input tokens and $2.15 per million output tokens; Qwen2.5-VL 72B Instruct lists at $2.80 and $8.40.
Which has the bigger context window?
DeepSeek-R1 does, with 164K tokens against 131K.
How many benchmarks do DeepSeek-R1 and Qwen2.5-VL 72B Instruct share?
1 benchmark has published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Qwen2.5-VL 72B Instruct has 6.