Model comparison
Claude 3 Opus vs DeepSeek-R1
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 29.5 on the Noometry Index.
Last verified . 33 shared benchmarks.
DeepSeekSummary
- They share 33 benchmarks with published results for both. Claude 3 Opus scores higher in 0 categories and DeepSeek-R1 in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-R1 leads 43.8 to 14.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 4.7% for Claude 3 Opus and 66.4% for DeepSeek-R1.
Side by side
| Claude 3 Opus | DeepSeek-R1 | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 29.5 | 42.3 |
| Released | 2024-02-29 | 2025-01-20 |
| Weights | Proprietary | Proprietary |
| Context window | — | 164K |
| Max output | — | 64K |
| Input $ / M tokens | — | $0.50 |
| Output $ / M tokens | — | $2.15 |
| Results tracked | 46 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-R1 leads
Claude 3 Opus: 32.9 (#267), DeepSeek-R1: 46.3 (#68)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| WeirdML | 19.2% | 41.6% |
| LiveBench Coding | 38.6% | 66.7% |
| LMArena Coding | 1264 | 1427 |
| Aider Polyglot | — | 71.4% |
| SciCode | — | 35.7% |
| BigCodeBench Instruct | 45.5% | — |
| BigCodeBench Complete | 57.4% | — |
| ALE-Bench | — | 804.12 |
| AlgoTune | — | 1.7 |
| HumanEval+ | 77.4% | — |
| MBPP+ | 73.3% | — |
Agentic & Tool Use DeepSeek-R1 leads
Claude 3 Opus: 24.6 (#116), DeepSeek-R1: 30.7 (#75)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| METR Time Horizons | 29.5% | 53.8% |
| Cybench | 10% | — |
| DeepResearch Bench | — | 35.1% |
| BALROG | — | 34.9% |
Reasoning DeepSeek-R1 leads
Claude 3 Opus: 14.6 (#324), DeepSeek-R1: 18.6 (#278)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| SimpleBench | 23.5% | 40.8% |
| LiveBench Reasoning | 40.6% | 83.2% |
| LMArena Hard Prompts | 1245 | 1416 |
| LiveBench Data Analysis | 57.9% | 69.8% |
| Epoch Capabilities Index | 126.91 | 141.29 |
| ForecastBench | 58.4 | 60 |
| LiveBench | 49.2% | 71.6% |
| ARC-AGI-2 | — | 1.3% |
| Kagi LLM Benchmark | — | 69.4% |
| ARC-AGI-1 | — | 21.2% |
| CritPt | — | 1.1% |
| Chess Puzzles | 5% | — |
| EnigmaEval | 0.8% | — |
| DTBench | 61.6% | — |
| LMCA | 17% | — |
| WinoGrande | 88.5% | — |
Math DeepSeek-R1 leads
Claude 3 Opus: 14.8 (#299), DeepSeek-R1: 43.8 (#79)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 4.7% | 66.4% |
| LiveBench Math | 43.6% | 80.7% |
| LMArena Math | 1273 | 1400 |
| MATH Level 5 | 37.5% | 96.6% |
| Omni-MATH | — | 42.4% |
Knowledge DeepSeek-R1 leads
Claude 3 Opus: 24.5 (#267), DeepSeek-R1: 44.5 (#87)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| GPQA Diamond | 47.2% | 76.3% |
| Confabulations | 22.7% | 12.7% |
| LMArena Expert | 1223 | 1394 |
| SimpleQA Verified | 12.6% | — |
| MMLU-Pro | — | 79.3% |
| Vectara Hallucination Rate | — | 11.3% |
| GPQA (HELM) | — | 66.6% |
| MMLU | 84.6% | — |
Multimodal Not comparable
Claude 3 Opus: 27.1 (#116), DeepSeek-R1: —
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| LMArena Vision | 1023 | — |
Multilingual DeepSeek-R1 leads
Claude 3 Opus: 41.4 (#207), DeepSeek-R1: 52.4 (#85)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| LMArena Non-English | 1258 | 1412 |
| LMArena Chinese | 1248 | 1442 |
| LMArena French | 1275 | 1417 |
| LMArena German | 1258 | 1404 |
| LMArena Japanese | 1204 | 1391 |
| LMArena Korean | 1187 | 1360 |
| LMArena Russian | 1280 | 1423 |
| LMArena Spanish | 1246 | 1411 |
Instruction Following DeepSeek-R1 leads
Claude 3 Opus: 64.1 (#228), DeepSeek-R1: 72.0 (#143)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| LiveBench Instruction Following | 63.9% | 80.5% |
| LMArena Instruction Following | 1248 | 1382 |
| IFEval | — | 78.4% |
Long Context DeepSeek-R1 leads
Claude 3 Opus: 38.2 (#202), DeepSeek-R1: 45.4 (#36)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| LMArena Longer Query | 1259 | 1391 |
| Fiction.LiveBench | — | 75% |
Writing & Preference DeepSeek-R1 leads
Claude 3 Opus: 47.2 (#213), DeepSeek-R1: 61.4 (#88)
| Benchmark | Claude 3 Opus | DeepSeek-R1 |
|---|---|---|
| LMArena Text | 1262 | 1428 |
| LMArena Creative Writing | 1235 | 1405 |
| LMArena Multi-Turn | 1275 | 1405 |
| LiveBench Language | 50.4% | 48.5% |
| Short-Story Creative Writing | — | 83% |
| EQ-Bench Creative Writing | — | 1500 |
| WildBench | — | 82.8% |
Frequently asked questions
Is Claude 3 Opus better than DeepSeek-R1?
DeepSeek-R1 is the stronger model overall, scoring 42.3 to 29.5 on the Noometry Index.
Is Claude 3 Opus or DeepSeek-R1 better for coding?
DeepSeek-R1 scores higher on coding benchmarks: 46.3 versus 32.9 in the Noometry coding category.
How many benchmarks do Claude 3 Opus and DeepSeek-R1 share?
33 benchmarks have published results for both models. Claude 3 Opus has 46 scored results on Noometry and DeepSeek-R1 has 52.