Model comparison
Claude 3.7 Sonnet vs DeepSeek-V3
Claude 3.7 Sonnet and DeepSeek-V3 score almost the same on the Noometry Index (39.5 vs 39.5), so choose on price, context window or the category you care about most.
Last verified . 42 shared benchmarks.
DeepSeekSummary
- They share 42 benchmarks with published results for both. Claude 3.7 Sonnet scores higher in 4 categories and DeepSeek-V3 in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in long context, where Claude 3.7 Sonnet leads 50.3 to 34.0.
- The biggest single-benchmark swing is Fiction.LiveBench: 83.3% for Claude 3.7 Sonnet and 50% for DeepSeek-V3.
- DeepSeek-V3 has downloadable open weights; the other is API-only.
Side by side
| Claude 3.7 Sonnet | DeepSeek-V3 | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 39.5 | 39.5 |
| Released | 2025-02-24 | 2024-12-26 |
| Weights | Proprietary | Open |
| Context window | — | 164K |
| Max output | — | 164K |
| Input $ / M tokens | — | $0.24 |
| Output $ / M tokens | — | $0.90 |
| Results tracked | 58 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3 leads
Claude 3.7 Sonnet: 40.6 (#136), DeepSeek-V3: 42.3 (#106)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| Aider Polyglot | 64.9% | 55.1% |
| LiveBench Coding | 74.5% | 70.9% |
| LMArena Coding | 1361 | 1368 |
| SWE-bench Verified | 61% | — |
| SWE-bench Verified (bash only) | 52.8% | — |
| SciCode | — | 35.8% |
| GSO | 3.8% | — |
| WeirdML | — | 36.1% |
| BigCodeBench Instruct | — | 50% |
| BigCodeBench Complete | — | 62.2% |
| CadEval | 54% | — |
| HumanEval+ | — | 86.6% |
| MBPP+ | — | 73% |
Agentic & Tool Use Not comparable
Claude 3.7 Sonnet: 34.1 (#50), DeepSeek-V3: —
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| METR Time Horizons | 60% | 49.6% |
| TheAgentCompany | 30.9% | — |
| Cybench | 20% | — |
| DeepResearch Bench | 43.6% | — |
| OSWorld | 35.8% | — |
Reasoning DeepSeek-V3 leads
Claude 3.7 Sonnet: 18.6 (#277), DeepSeek-V3: 20.5 (#236)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| SimpleBench | 46.4% | 27.2% |
| LiveBench Reasoning | 87.8% | 65.8% |
| LMArena Hard Prompts | 1333 | 1365 |
| LiveBench Data Analysis | 74% | 60.9% |
| Epoch Capabilities Index | 141.16 | 135.94 |
| ForecastBench | 61.8 | 59.1 |
| LiveBench | 76.1% | 66.9% |
| ARC-AGI-2 | 0.9% | — |
| Kagi LLM Benchmark | — | 52.3% |
| ARC-AGI-1 | 28.6% | — |
| CritPt | — | 0% |
| EnigmaEval | 4.2% | — |
| DTBench | — | 64.8% |
| LMCA | — | 15.5% |
| BIG-Bench Hard | — | 87.5% |
| HellaSwag | — | 88.9% |
| PIQA | — | 84.7% |
| WinoGrande | — | 85.2% |
Math Claude 3.7 Sonnet leads
Claude 3.7 Sonnet: 37.5 (#153), DeepSeek-V3: 32.1 (#219)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 57.8% | 37.8% |
| Omni-MATH | 33% | 40.3% |
| LiveBench Math | 79% | 73.5% |
| LMArena Math | 1337 | 1373 |
| MATH Level 5 | 91.2% | 75.5% |
| FrontierMath (Feb 2025 set) | 4.1% | 1.7% |
Knowledge Claude 3.7 Sonnet leads
Claude 3.7 Sonnet: 39.8 (#130), DeepSeek-V3: 37.5 (#155)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| GPQA Diamond | 79.7% | 67.6% |
| MMLU-Pro | 78.4% | 72.3% |
| Confabulations | 14.7% | 26.1% |
| GPQA (HELM) | 60.8% | 53.8% |
| LMArena Expert | 1321 | 1351 |
| Humanity's Last Exam | 8% | — |
| Vectara Hallucination Rate | — | 6.1% |
| ARC (AI2) Challenge | — | 95.3% |
| MMLU | — | 87.2% |
| TriviaQA | — | 82.9% |
Multimodal Not comparable
Claude 3.7 Sonnet: 33.7 (#95), DeepSeek-V3: —
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Vision | 1169 | — |
| GeoBench | 68% | — |
| VPCT | 39% | — |
| SpatialViz-Bench | 33.9% | — |
Multilingual DeepSeek-V3 leads
Claude 3.7 Sonnet: 44.1 (#179), DeepSeek-V3: 48.5 (#143)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Non-English | 1296 | 1358 |
| LMArena Chinese | 1299 | 1391 |
| LMArena French | 1303 | 1385 |
| LMArena German | 1301 | 1374 |
| LMArena Japanese | 1267 | 1333 |
| LMArena Korean | 1249 | 1319 |
| LMArena Russian | 1311 | 1373 |
| LMArena Spanish | 1298 | 1358 |
Instruction Following Too close to call
Claude 3.7 Sonnet: 72.9 (#125), DeepSeek-V3: 72.8 (#130)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| LiveBench Instruction Following | 81.3% | 81.5% |
| IFEval | 83.4% | 83.2% |
| LMArena Instruction Following | 1352 | 1345 |
Long Context Claude 3.7 Sonnet leads
Claude 3.7 Sonnet: 50.3 (#10), DeepSeek-V3: 34.0 (#253)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| Fiction.LiveBench | 83.3% | 50% |
| LMArena Longer Query | 1373 | 1352 |
Writing & Preference DeepSeek-V3 leads
Claude 3.7 Sonnet: 54.4 (#150), DeepSeek-V3: 57.4 (#130)
| Benchmark | Claude 3.7 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Text | 1314 | 1375 |
| LMArena Creative Writing | 1332 | 1364 |
| Short-Story Creative Writing | 81.1% | 77% |
| EQ-Bench Creative Writing | 1412 | 1472 |
| WildBench | 81.4% | 83% |
| LMArena Multi-Turn | 1339 | 1389 |
| LiveBench Language | 59.9% | 49.1% |
Frequently asked questions
Is Claude 3.7 Sonnet better than DeepSeek-V3?
Claude 3.7 Sonnet and DeepSeek-V3 score almost the same on the Noometry Index (39.5 vs 39.5), so choose on price, context window or the category you care about most.
Is Claude 3.7 Sonnet or DeepSeek-V3 better for coding?
DeepSeek-V3 scores higher on coding benchmarks: 42.3 versus 40.6 in the Noometry coding category.
How many benchmarks do Claude 3.7 Sonnet and DeepSeek-V3 share?
42 benchmarks have published results for both models. Claude 3.7 Sonnet has 58 scored results on Noometry and DeepSeek-V3 has 60.