Model comparison
Claude 3.5 Sonnet vs DeepSeek-V3
DeepSeek-V3 is the stronger model overall, scoring 39.5 to 34.6 on the Noometry Index.
Last verified . 48 shared benchmarks.
DeepSeekSummary
- They share 48 benchmarks with published results for both. Claude 3.5 Sonnet scores higher in 2 categories and DeepSeek-V3 in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-V3 leads 32.1 to 19.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 8.5% for Claude 3.5 Sonnet and 37.8% for DeepSeek-V3.
- DeepSeek-V3 has downloadable open weights; the other is API-only.
Side by side
| Claude 3.5 Sonnet | DeepSeek-V3 | |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 34.6 | 39.5 |
| Released | 2024-06-20 | 2024-12-26 |
| Weights | Proprietary | Open |
| Context window | — | 164K |
| Max output | — | 164K |
| Input $ / M tokens | — | $0.24 |
| Output $ / M tokens | — | $0.90 |
| Results tracked | 60 | 60 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3 leads
Claude 3.5 Sonnet: 39.0 (#165), DeepSeek-V3: 42.3 (#106)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| Aider Polyglot | 51.6% | 55.1% |
| WeirdML | 40% | 36.1% |
| BigCodeBench Instruct | 46.8% | 50% |
| LiveBench Coding | 67.1% | 70.9% |
| LMArena Coding | 1342 | 1368 |
| BigCodeBench Complete | 58.6% | 62.2% |
| HumanEval+ | 81.7% | 86.6% |
| MBPP+ | 74.3% | 73% |
| SciCode | — | 35.8% |
| GSO | 4.6% | — |
| CadEval | 48% | — |
Agentic & Tool Use Not comparable
Claude 3.5 Sonnet: 32.3 (#67), DeepSeek-V3: —
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| METR Time Horizons | 45.2% | 49.6% |
| TheAgentCompany | 24% | — |
| Cybench | 17.5% | — |
| BALROG | 32.6% | — |
Reasoning Claude 3.5 Sonnet leads
Claude 3.5 Sonnet: 23.1 (#183), DeepSeek-V3: 20.5 (#236)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| SimpleBench | 41.4% | 27.2% |
| LiveBench Reasoning | 56.7% | 65.8% |
| LMArena Hard Prompts | 1305 | 1365 |
| DTBench | 67.8% | 64.8% |
| LiveBench Data Analysis | 55% | 60.9% |
| Epoch Capabilities Index | 133.55 | 135.94 |
| ForecastBench | 60.7 | 59.1 |
| LiveBench | 59% | 66.9% |
| Kagi LLM Benchmark | — | 52.3% |
| CritPt | — | 0% |
| EnigmaEval | 0.9% | — |
| LMCA | — | 15.5% |
| BIG-Bench Hard | — | 87.5% |
| HellaSwag | — | 88.9% |
| PIQA | — | 84.7% |
| WinoGrande | — | 85.2% |
Math DeepSeek-V3 leads
Claude 3.5 Sonnet: 19.2 (#288), DeepSeek-V3: 32.1 (#219)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.5% | 37.8% |
| Omni-MATH | 27.6% | 40.3% |
| LiveBench Math | 52.3% | 73.5% |
| LMArena Math | 1307 | 1373 |
| MATH Level 5 | 56.9% | 75.5% |
| FrontierMath (Feb 2025 set) | 2.1% | 1.7% |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge DeepSeek-V3 leads
Claude 3.5 Sonnet: 28.6 (#245), DeepSeek-V3: 37.5 (#155)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| GPQA Diamond | 55.3% | 67.6% |
| MMLU-Pro | 77.7% | 72.3% |
| Confabulations | 19.9% | 26.1% |
| GPQA (HELM) | 56.5% | 53.8% |
| LMArena Expert | 1265 | 1351 |
| MMLU | 87.3% | 87.2% |
| Humanity's Last Exam | 4.1% | — |
| Vectara Hallucination Rate | — | 6.1% |
| ARC (AI2) Challenge | — | 95.3% |
| TriviaQA | — | 82.9% |
Multimodal Not comparable
Claude 3.5 Sonnet: 26.5 (#120), DeepSeek-V3: —
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Vision | 1125 | — |
| Video-MME | 60% | — |
| GeoBench | 62% | — |
| VPCT | 33% | — |
Multilingual DeepSeek-V3 leads
Claude 3.5 Sonnet: 43.2 (#185), DeepSeek-V3: 48.5 (#143)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Non-English | 1283 | 1358 |
| LMArena Chinese | 1272 | 1391 |
| LMArena French | 1305 | 1385 |
| LMArena German | 1297 | 1374 |
| LMArena Japanese | 1234 | 1333 |
| LMArena Korean | 1200 | 1319 |
| LMArena Russian | 1306 | 1373 |
| LMArena Spanish | 1290 | 1358 |
Instruction Following DeepSeek-V3 leads
Claude 3.5 Sonnet: 68.8 (#182), DeepSeek-V3: 72.8 (#130)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| LiveBench Instruction Following | 69.3% | 81.5% |
| IFEval | 85.5% | 83.2% |
| LMArena Instruction Following | 1297 | 1345 |
Long Context Claude 3.5 Sonnet leads
Claude 3.5 Sonnet: 39.9 (#167), DeepSeek-V3: 34.0 (#253)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Longer Query | 1311 | 1352 |
| Fiction.LiveBench | — | 50% |
Writing & Preference DeepSeek-V3 leads
Claude 3.5 Sonnet: 52.9 (#164), DeepSeek-V3: 57.4 (#130)
| Benchmark | Claude 3.5 Sonnet | DeepSeek-V3 |
|---|---|---|
| LMArena Text | 1298 | 1375 |
| LMArena Creative Writing | 1292 | 1364 |
| Short-Story Creative Writing | 80.3% | 77% |
| EQ-Bench Creative Writing | 1451 | 1472 |
| WildBench | 79.2% | 83% |
| LMArena Multi-Turn | 1326 | 1389 |
| LiveBench Language | 53.8% | 49.1% |
Frequently asked questions
Is Claude 3.5 Sonnet better than DeepSeek-V3?
DeepSeek-V3 is the stronger model overall, scoring 39.5 to 34.6 on the Noometry Index.
Is Claude 3.5 Sonnet or DeepSeek-V3 better for coding?
DeepSeek-V3 scores higher on coding benchmarks: 42.3 versus 39.0 in the Noometry coding category.
How many benchmarks do Claude 3.5 Sonnet and DeepSeek-V3 share?
48 benchmarks have published results for both models. Claude 3.5 Sonnet has 60 scored results on Noometry and DeepSeek-V3 has 60.