# Deepseek Coder v2 vs Llama 3.1-405B

> Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 30.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/deepseek-coder-v2-vs-llama-3-1-405b
- Last updated: 2026-10-11
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. Deepseek Coder v2 scores higher in 4 categories and Llama 3.1-405B in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Deepseek Coder v2 leads 34.9 to 18.4.

## Snapshot

| | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| Provider | DeepSeek | Meta |
| Noometry Index | 35.9 | 30.7 |
| Rank | 220 | 288 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Open |

## Coding

- Deepseek Coder v2: 38.1 (#183)
- Llama 3.1-405B: 33.1 (#262)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Coding | 1251 | 1291 |
| WeirdML | — | 21.4% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |

## Agentic & Tool Use

- Deepseek Coder v2: —
- Llama 3.1-405B: 21.0 (#140)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| TheAgentCompany | — | 7.4% |
| Cybench | — | 7.5% |

## Reasoning

- Deepseek Coder v2: 23.6 (#176)
- Llama 3.1-405B: 16.8 (#300)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1269 |
| WinoGrande | 83.7% | 89.2% |
| SimpleBench | — | 23% |
| Kagi LLM Benchmark | — | 45% |
| DTBench | — | 61.4% |
| BIG-Bench Hard | — | 82.9% |
| Epoch Capabilities Index | — | 128.75 |
| ForecastBench | — | 59.9 |
| HellaSwag | — | 89.2% |
| PIQA | — | 85.9% |

## Math

- Deepseek Coder v2: 34.9 (#190)
- Llama 3.1-405B: 18.4 (#290)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Math | 1241 | 1281 |
| OTIS Mock AIME 2024-2025 | — | 9.7% |
| Omni-MATH | — | 24.9% |
| MATH Level 5 | — | 49.8% |
| GSM8K | 94.5% | — |

## Knowledge

- Deepseek Coder v2: 32.3 (#212)
- Llama 3.1-405B: 30.4 (#227)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Expert | 1181 | 1243 |
| ARC (AI2) Challenge | 64.3% | 95.3% |
| GPQA Diamond | — | 50.9% |
| MMLU-Pro | — | 72.3% |
| Confabulations | — | 17.6% |
| GPQA (HELM) | — | 52.2% |
| MMLU | — | 84.5% |
| TriviaQA | — | 82.7% |

## Multilingual

- Deepseek Coder v2: 36.3 (#240)
- Llama 3.1-405B: 40.7 (#214)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Non-English | 1182 | 1248 |
| LMArena Chinese | 1201 | 1242 |
| LMArena French | 1185 | 1279 |
| LMArena German | 1164 | 1252 |
| LMArena Japanese | 1126 | 1208 |
| LMArena Korean | 1104 | 1184 |
| LMArena Russian | 1188 | 1265 |
| LMArena Spanish | 1153 | 1260 |

## Instruction Following

- Deepseek Coder v2: 61.7 (#242)
- Llama 3.1-405B: 65.9 (#214)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Instruction Following | 1180 | 1259 |
| IFEval | — | 81.1% |

## Long Context

- Deepseek Coder v2: 37.0 (#224)
- Llama 3.1-405B: 38.4 (#197)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Longer Query | 1219 | 1266 |

## Writing & Preference

- Deepseek Coder v2: 38.2 (#253)
- Llama 3.1-405B: 38.9 (#251)

| Benchmark | Deepseek Coder v2 | Llama 3.1-405B |
|---|---|---|
| LMArena Text | 1191 | 1284 |
| LMArena Creative Writing | 1120 | 1262 |
| LMArena Multi-Turn | 1177 | 1297 |
| EQ-Bench Creative Writing | — | 870 |
| WildBench | — | 78.3% |

## FAQ

### Is Deepseek Coder v2 better than Llama 3.1-405B?

Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 30.7 on the Noometry Index.

### Is Deepseek Coder v2 or Llama 3.1-405B better for coding?

Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 33.1 in the Noometry coding category.

### How many benchmarks do Deepseek Coder v2 and Llama 3.1-405B share?

19 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and Llama 3.1-405B has 42.