Model comparison
DeepSeek-V3.1 vs Llama 3.1-405B
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 30.7 on the Noometry Index.
Last verified . 24 shared benchmarks.
DeepSeek
MetaSummary
- They share 24 benchmarks with published results for both. DeepSeek-V3.1 scores higher in 7 categories and Llama 3.1-405B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.1 leads 60.3 to 38.9.
- The biggest single-benchmark swing is DTBench: 82.7% for DeepSeek-V3.1 and 61.4% for Llama 3.1-405B.
Side by side
| DeepSeek-V3.1 | Llama 3.1-405B | |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 42.8 | 30.7 |
| Released | 2025-08-21 | 2024-07-23 |
| Weights | Open | Open |
| Context window | 164K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.25 | — |
| Output $ / M tokens | $0.95 | — |
| Results tracked | 27 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1 leads
DeepSeek-V3.1: 40.3 (#144), Llama 3.1-405B: 33.1 (#262)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| WeirdML | 38.4% | 21.4% |
| LMArena Coding | 1417 | 1291 |
Agentic & Tool Use Not comparable
DeepSeek-V3.1: —, Llama 3.1-405B: 21.0 (#140)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| TheAgentCompany | — | 7.4% |
| Cybench | — | 7.5% |
Reasoning DeepSeek-V3.1 leads
DeepSeek-V3.1: 27.9 (#110), Llama 3.1-405B: 16.8 (#300)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| SimpleBench | 40% | 23% |
| Kagi LLM Benchmark | 53.2% | 45% |
| LMArena Hard Prompts | 1417 | 1269 |
| DTBench | 82.7% | 61.4% |
| Epoch Capabilities Index | 139.92 | 128.75 |
| ForecastBench | 58 | 59.9 |
| LMCA | 24.3% | — |
| BIG-Bench Hard | — | 82.9% |
| HellaSwag | — | 89.2% |
| PIQA | — | 85.9% |
| WinoGrande | — | 89.2% |
Math DeepSeek-V3.1 leads
DeepSeek-V3.1: 38.9 (#122), Llama 3.1-405B: 18.4 (#290)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| LMArena Math | 1420 | 1281 |
| OTIS Mock AIME 2024-2025 | — | 9.7% |
| Omni-MATH | — | 24.9% |
| MATH Level 5 | — | 49.8% |
Knowledge DeepSeek-V3.1 leads
DeepSeek-V3.1: 43.7 (#90), Llama 3.1-405B: 30.4 (#227)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| LMArena Expert | 1405 | 1243 |
| GPQA Diamond | — | 50.9% |
| MMLU-Pro | — | 72.3% |
| Confabulations | — | 17.6% |
| Vectara Hallucination Rate | 5.5% | — |
| GPQA (HELM) | — | 52.2% |
| ARC (AI2) Challenge | — | 95.3% |
| MMLU | — | 84.5% |
| TriviaQA | — | 82.7% |
Multilingual DeepSeek-V3.1 leads
DeepSeek-V3.1: 51.6 (#106), Llama 3.1-405B: 40.7 (#214)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| LMArena Non-English | 1400 | 1248 |
| LMArena Chinese | 1469 | 1242 |
| LMArena French | 1447 | 1279 |
| LMArena German | 1411 | 1252 |
| LMArena Japanese | 1378 | 1208 |
| LMArena Korean | 1337 | 1184 |
| LMArena Russian | 1405 | 1265 |
| LMArena Spanish | 1431 | 1260 |
Instruction Following DeepSeek-V3.1 leads
DeepSeek-V3.1: 73.9 (#110), Llama 3.1-405B: 65.9 (#214)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| LMArena Instruction Following | 1400 | 1259 |
| IFEval | — | 81.1% |
Long Context Llama 3.1-405B leads
DeepSeek-V3.1: 36.3 (#232), Llama 3.1-405B: 38.4 (#197)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| LMArena Longer Query | 1422 | 1266 |
| Fiction.LiveBench | 52.8% | — |
Writing & Preference DeepSeek-V3.1 leads
DeepSeek-V3.1: 60.3 (#98), Llama 3.1-405B: 38.9 (#251)
| Benchmark | DeepSeek-V3.1 | Llama 3.1-405B |
|---|---|---|
| LMArena Text | 1420 | 1284 |
| LMArena Creative Writing | 1401 | 1262 |
| EQ-Bench Creative Writing | 1436 | 870 |
| LMArena Multi-Turn | 1408 | 1297 |
| WildBench | — | 78.3% |
Frequently asked questions
Is DeepSeek-V3.1 better than Llama 3.1-405B?
DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 30.7 on the Noometry Index.
Is DeepSeek-V3.1 or Llama 3.1-405B better for coding?
DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 33.1 in the Noometry coding category.
How many benchmarks do DeepSeek-V3.1 and Llama 3.1-405B share?
24 benchmarks have published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Llama 3.1-405B has 42.