# Claude 3 Sonnet vs Deepseek Coder v2

> Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 29.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-sonnet-vs-deepseek-coder-v2
- Last updated: 2026-10-10
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. Claude 3 Sonnet scores higher in 3 categories and Deepseek Coder v2 in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Deepseek Coder v2 leads 34.9 to 10.7.
- The biggest single-benchmark swing is BigCodeBench Complete: 53.8% for Claude 3 Sonnet and 59.7% for Deepseek Coder v2.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 29.0 | 35.9 |
| Rank | 319 | 220 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Sonnet: 29.6 (#302)
- Deepseek Coder v2: 38.1 (#183)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| BigCodeBench Instruct | 42.7% | 48.2% |
| LMArena Coding | 1223 | 1251 |
| BigCodeBench Complete | 53.8% | 59.7% |
| HumanEval+ | 64% | 82.3% |
| MBPP+ | 69.3% | 75.1% |
| WeirdML | 10.2% | — |

## Reasoning

- Claude 3 Sonnet: 20.5 (#237)
- Deepseek Coder v2: 23.6 (#176)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Hard Prompts | 1197 | 1207 |
| WinoGrande | 75.1% | 83.7% |
| DTBench | 53.6% | — |
| Epoch Capabilities Index | 120.7 | — |

## Math

- Claude 3 Sonnet: 10.7 (#310)
- Deepseek Coder v2: 34.9 (#190)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Math | 1213 | 1241 |
| OTIS Mock AIME 2024-2025 | 2.5% | — |
| MATH Level 5 | 18.2% | — |
| GSM8K | — | 94.5% |

## Knowledge

- Claude 3 Sonnet: 21.1 (#276)
- Deepseek Coder v2: 32.3 (#212)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Expert | 1173 | 1181 |
| GPQA Diamond | 40.6% | — |
| ARC (AI2) Challenge | — | 64.3% |
| MMLU | 75.9% | — |

## Multimodal

- Claude 3 Sonnet: 25.2 (#125)
- Deepseek Coder v2: —

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Vision | 984 | — |

## Multilingual

- Claude 3 Sonnet: 37.8 (#234)
- Deepseek Coder v2: 36.3 (#240)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Non-English | 1205 | 1182 |
| LMArena Chinese | 1189 | 1201 |
| LMArena French | 1229 | 1185 |
| LMArena German | 1204 | 1164 |
| LMArena Japanese | 1131 | 1126 |
| LMArena Korean | 1128 | 1104 |
| LMArena Russian | 1227 | 1188 |
| LMArena Spanish | 1204 | 1153 |

## Instruction Following

- Claude 3 Sonnet: 62.8 (#235)
- Deepseek Coder v2: 61.7 (#242)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Instruction Following | 1199 | 1180 |

## Long Context

- Claude 3 Sonnet: 36.7 (#228)
- Deepseek Coder v2: 37.0 (#224)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Longer Query | 1211 | 1219 |

## Writing & Preference

- Claude 3 Sonnet: 42.1 (#238)
- Deepseek Coder v2: 38.2 (#253)

| Benchmark | Claude 3 Sonnet | Deepseek Coder v2 |
|---|---|---|
| LMArena Text | 1218 | 1191 |
| LMArena Creative Writing | 1186 | 1120 |
| LMArena Multi-Turn | 1227 | 1177 |

## FAQ

### Is Claude 3 Sonnet better than Deepseek Coder v2?

Deepseek Coder v2 is the stronger model overall, scoring 35.9 to 29.0 on the Noometry Index.

### Is Claude 3 Sonnet or Deepseek Coder v2 better for coding?

Deepseek Coder v2 scores higher on coding benchmarks: 38.1 versus 29.6 in the Noometry coding category.

### How many benchmarks do Claude 3 Sonnet and Deepseek Coder v2 share?

22 benchmarks have published results for both models. Claude 3 Sonnet has 30 scored results on Noometry and Deepseek Coder v2 has 24.