# Claude 3 Sonnet vs DeepSeek-V2.5 (Sep 2024)

> DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 28.8 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-sonnet-vs-deepseek-v2-5
- Last updated: 2026-10-10
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Claude 3 Sonnet scores higher in 0 categories and DeepSeek-V2.5 (Sep 2024) in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek-V2.5 (Sep 2024) leads 33.3 to 10.0.
- The biggest single-benchmark swing is BigCodeBench Instruct: 42.7% for Claude 3 Sonnet and 48.6% for DeepSeek-V2.5 (Sep 2024).
- DeepSeek-V2.5 (Sep 2024) has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 28.8 | 37.0 |
| Rank | 311 | 197 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Sonnet: 29.2 (#296)
- DeepSeek-V2.5 (Sep 2024): 31.3 (#274)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| BigCodeBench Instruct | 42.7% | 48.6% |
| LMArena Coding | 1223 | 1310 |
| BigCodeBench Complete | 53.8% | 53.2% |
| HumanEval+ | 64% | 83.5% |
| MBPP+ | 69.3% | 74.1% |
| Aider Polyglot | — | 17.8% |
| WeirdML | 10.2% | — |

## Reasoning

- Claude 3 Sonnet: 19.7 (#232)
- DeepSeek-V2.5 (Sep 2024): 24.5 (#145)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Hard Prompts | 1197 | 1289 |
| DTBench | 53.6% | — |
| Epoch Capabilities Index | 120.69 | — |
| WinoGrande | 75.1% | — |

## Math

- Claude 3 Sonnet: 10.0 (#305)
- DeepSeek-V2.5 (Sep 2024): 33.3 (#177)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Math | 1213 | 1289 |
| OTIS Mock AIME 2024-2025 | 2.5% | — |
| MATH Level 5 | 18.2% | — |

## Knowledge

- Claude 3 Sonnet: 21.1 (#271)
- DeepSeek-V2.5 (Sep 2024): 34.2 (#190)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Expert | 1173 | 1266 |
| GPQA Diamond | 40.6% | — |
| MMLU | 75.9% | — |

## Multimodal

- Claude 3 Sonnet: 24.7 (#120)
- DeepSeek-V2.5 (Sep 2024): —

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Vision | 984 | — |

## Multilingual

- Claude 3 Sonnet: 38.0 (#229)
- DeepSeek-V2.5 (Sep 2024): 42.7 (#188)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Non-English | 1205 | 1273 |
| LMArena Chinese | 1188 | 1317 |
| LMArena French | 1228 | 1288 |
| LMArena German | 1204 | 1258 |
| LMArena Japanese | 1131 | 1228 |
| LMArena Korean | 1128 | 1209 |
| LMArena Russian | 1227 | 1289 |
| LMArena Spanish | 1204 | 1248 |

## Instruction Following

- Claude 3 Sonnet: 63.1 (#230)
- DeepSeek-V2.5 (Sep 2024): 67.7 (#189)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Instruction Following | 1200 | 1280 |

## Long Context

- Claude 3 Sonnet: 36.7 (#223)
- DeepSeek-V2.5 (Sep 2024): 39.5 (#169)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Longer Query | 1211 | 1301 |

## Writing & Preference

- Claude 3 Sonnet: 42.3 (#233)
- DeepSeek-V2.5 (Sep 2024): 50.0 (#182)

| Benchmark | Claude 3 Sonnet | DeepSeek-V2.5 (Sep 2024) |
|---|---|---|
| LMArena Text | 1218 | 1294 |
| LMArena Creative Writing | 1186 | 1285 |
| LMArena Multi-Turn | 1227 | 1297 |

## FAQ

### Is Claude 3 Sonnet better than DeepSeek-V2.5 (Sep 2024)?

DeepSeek-V2.5 (Sep 2024) is the stronger model overall, scoring 37.0 to 28.8 on the Noometry Index.

### Is Claude 3 Sonnet or DeepSeek-V2.5 (Sep 2024) better for coding?

DeepSeek-V2.5 (Sep 2024) scores higher on coding benchmarks: 31.3 versus 29.2 in the Noometry coding category.

### How many benchmarks do Claude 3 Sonnet and DeepSeek-V2.5 (Sep 2024) share?

21 benchmarks have published results for both models. Claude 3 Sonnet has 30 scored results on Noometry and DeepSeek-V2.5 (Sep 2024) has 22.