Model comparison
DeepSeek-R1 vs o1-mini
DeepSeek-R1 is the stronger model overall, scoring 42.0 to 33.7 on the Noometry Index.
Last verified . 35 shared benchmarks.
DeepSeekSummary
- They share 35 benchmarks with published results for both. DeepSeek-R1 scores higher in 9 categories and o1-mini in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-R1 leads 61.5 to 48.6.
- The biggest single-benchmark swing is Aider Polyglot: 71.4% for DeepSeek-R1 and 32.9% for o1-mini.
Side by side
| DeepSeek-R1 | o1-mini | |
|---|---|---|
| Provider | DeepSeek | OpenAI |
| Noometry Index | 42.0 | 33.7 |
| Released | 2025-01-20 | 2024-09-12 |
| Weights | Proprietary | Proprietary |
| Context window | 164K | — |
| Max output | 64K | — |
| Input $ / M tokens | $0.50 | — |
| Output $ / M tokens | $2.15 | — |
| Results tracked | 52 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1 leads
DeepSeek-R1: 45.9 (#63), o1-mini: 35.1 (#218)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| Aider Polyglot | 71.4% | 32.9% |
| WeirdML | 41.6% | 36.3% |
| LiveBench Coding | 66.7% | 48% |
| LMArena Coding | 1427 | 1362 |
| SciCode | 35.7% | — |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 78.8% |
Agentic & Tool Use DeepSeek-R1 leads
DeepSeek-R1: 30.4 (#68), o1-mini: 24.3 (#112)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| Cybench | — | 10% |
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| METR Time Horizons | 53.8% | — |
Reasoning DeepSeek-R1 leads
DeepSeek-R1: 18.7 (#253), o1-mini: 8.9 (#340)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| ARC-AGI-2 | 1.3% | 0.8% |
| SimpleBench | 40.8% | 18.1% |
| ARC-AGI-1 | 21.2% | 14% |
| LiveBench Reasoning | 83.2% | 72.3% |
| LMArena Hard Prompts | 1415 | 1333 |
| LiveBench Data Analysis | 69.8% | 57.9% |
| Epoch Capabilities Index | 141.3 | 135.82 |
| LiveBench | 71.6% | 57.8% |
| Kagi LLM Benchmark | 69.4% | — |
| CritPt | 1.1% | — |
| ForecastBench | 60 | — |
Math DeepSeek-R1 leads
DeepSeek-R1: 42.0 (#70), o1-mini: 33.7 (#168)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 46.9% |
| LiveBench Math | 80.7% | 62% |
| LMArena Math | 1401 | 1358 |
| MATH Level 5 | 96.6% | 89.2% |
| Omni-MATH | 42.4% | — |
| FrontierMath (Feb 2025 set) | — | 1.7% |
Knowledge DeepSeek-R1 leads
DeepSeek-R1: 44.3 (#81), o1-mini: 34.8 (#180)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| GPQA Diamond | 76.3% | 62.4% |
| Confabulations | 12.7% | 18.6% |
| LMArena Expert | 1394 | 1316 |
| MMLU-Pro | 79.3% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
Multilingual DeepSeek-R1 leads
DeepSeek-R1: 52.6 (#80), o1-mini: 43.8 (#177)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| LMArena Non-English | 1412 | 1289 |
| LMArena Chinese | 1443 | 1314 |
| LMArena French | 1415 | 1292 |
| LMArena German | 1405 | 1278 |
| LMArena Japanese | 1394 | 1245 |
| LMArena Korean | 1359 | 1223 |
| LMArena Russian | 1423 | 1283 |
| LMArena Spanish | 1410 | 1302 |
Instruction Following DeepSeek-R1 leads
DeepSeek-R1: 72.2 (#138), o1-mini: 67.0 (#200)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| LiveBench Instruction Following | 80.5% | 65.4% |
| LMArena Instruction Following | 1382 | 1304 |
| IFEval | 78.4% | — |
Long Context DeepSeek-R1 leads
DeepSeek-R1: 45.4 (#36), o1-mini: 40.1 (#156)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| LMArena Longer Query | 1391 | 1321 |
| Fiction.LiveBench | 75% | — |
Writing & Preference DeepSeek-R1 leads
DeepSeek-R1: 61.5 (#84), o1-mini: 48.6 (#196)
| Benchmark | DeepSeek-R1 | o1-mini |
|---|---|---|
| LMArena Text | 1428 | 1317 |
| LMArena Creative Writing | 1406 | 1244 |
| Short-Story Creative Writing | 83% | 64.9% |
| LMArena Multi-Turn | 1404 | 1314 |
| LiveBench Language | 48.5% | 40.9% |
| EQ-Bench Creative Writing | 1500 | — |
| WildBench | 82.8% | — |
Frequently asked questions
Is DeepSeek-R1 better than o1-mini?
DeepSeek-R1 is the stronger model overall, scoring 42.0 to 33.7 on the Noometry Index.
Is DeepSeek-R1 or o1-mini better for coding?
DeepSeek-R1 scores higher on coding benchmarks: 45.9 versus 35.1 in the Noometry coding category.
How many benchmarks do DeepSeek-R1 and o1-mini share?
35 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and o1-mini has 39.