# Claude 3.5 Haiku vs DeepSeek-V3.1

> DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 29.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-5-haiku-vs-deepseek-v3-1
- Last updated: 2026-10-12
- Shared benchmarks: 21

## Summary

- They share 21 benchmarks with published results for both. Claude 3.5 Haiku scores higher in 1 category and DeepSeek-V3.1 in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where DeepSeek-V3.1 leads 43.7 to 18.7.
- The biggest single-benchmark swing is DTBench: 56.7% for Claude 3.5 Haiku and 82.7% for DeepSeek-V3.1.
- DeepSeek-V3.1 has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 29.2 | 42.8 |
| Rank | 315 | 108 |
| Context | — | 164K |
| Input $/M | — | $0.25 |
| Output $/M | — | $0.95 |
| Weights | Proprietary | Open |

## Coding

- Claude 3.5 Haiku: 32.9 (#265)
- DeepSeek-V3.1: 40.3 (#144)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| WeirdML | 30.7% | 38.4% |
| LMArena Coding | 1286 | 1417 |
| Aider Polyglot | 28% | — |
| SciCode | 27.4% | — |
| BigCodeBench Instruct | 46.1% | — |
| LiveBench Coding | 51.4% | — |
| BigCodeBench Complete | 59% | — |
| CadEval | 32% | — |

## Agentic & Tool Use

- Claude 3.5 Haiku: 28.0 (#95)
- DeepSeek-V3.1: —

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| BALROG | 19.3% | — |

## Reasoning

- Claude 3.5 Haiku: 17.7 (#290)
- DeepSeek-V3.1: 27.9 (#110)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Hard Prompts | 1251 | 1417 |
| DTBench | 56.7% | 82.7% |
| Epoch Capabilities Index | 127.15 | 139.92 |
| SimpleBench | — | 40% |
| Kagi LLM Benchmark | — | 53.2% |
| CritPt | 0% | — |
| LiveBench Reasoning | 28.1% | — |
| LiveBench Data Analysis | 48.5% | — |
| LMCA | — | 24.3% |
| ForecastBench | — | 58 |
| LiveBench | 43.5% | — |

## Math

- Claude 3.5 Haiku: 14.7 (#300)
- DeepSeek-V3.1: 38.9 (#122)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Math | 1244 | 1420 |
| OTIS Mock AIME 2024-2025 | 4.3% | — |
| Omni-MATH | 22.4% | — |
| LiveBench Math | 35.5% | — |
| MATH Level 5 | 46.4% | — |
| FrontierMath (Feb 2025 set) | 0.3% | — |

## Knowledge

- Claude 3.5 Haiku: 18.7 (#281)
- DeepSeek-V3.1: 43.7 (#90)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Expert | 1208 | 1405 |
| GPQA Diamond | 38.1% | — |
| MMLU-Pro | 60.5% | — |
| Confabulations | 36.7% | — |
| Vectara Hallucination Rate | — | 5.5% |
| GPQA (HELM) | 36.3% | — |
| MMLU | 74.3% | — |

## Multimodal

- Claude 3.5 Haiku: 26.8 (#117)
- DeepSeek-V3.1: —

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Vision | 1092 | — |
| GeoBench | 34% | — |

## Multilingual

- Claude 3.5 Haiku: 40.0 (#218)
- DeepSeek-V3.1: 51.6 (#106)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Non-English | 1238 | 1400 |
| LMArena Chinese | 1229 | 1469 |
| LMArena French | 1264 | 1447 |
| LMArena German | 1237 | 1411 |
| LMArena Japanese | 1175 | 1378 |
| LMArena Korean | 1173 | 1337 |
| LMArena Russian | 1253 | 1405 |
| LMArena Spanish | 1261 | 1431 |

## Instruction Following

- Claude 3.5 Haiku: 62.9 (#234)
- DeepSeek-V3.1: 73.9 (#110)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Instruction Following | 1241 | 1400 |
| LiveBench Instruction Following | 61.9% | — |
| IFEval | 79.2% | — |

## Long Context

- Claude 3.5 Haiku: 38.3 (#200)
- DeepSeek-V3.1: 36.3 (#232)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Longer Query | 1261 | 1422 |
| Fiction.LiveBench | — | 52.8% |

## Writing & Preference

- Claude 3.5 Haiku: 42.7 (#234)
- DeepSeek-V3.1: 60.3 (#98)

| Benchmark | Claude 3.5 Haiku | DeepSeek-V3.1 |
|---|---|---|
| LMArena Text | 1255 | 1420 |
| LMArena Creative Writing | 1233 | 1401 |
| EQ-Bench Creative Writing | 1146 | 1436 |
| LMArena Multi-Turn | 1265 | 1408 |
| Short-Story Creative Writing | 73.5% | — |
| WildBench | 76% | — |
| LiveBench Language | 35.4% | — |

## FAQ

### Is Claude 3.5 Haiku better than DeepSeek-V3.1?

DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 29.2 on the Noometry Index.

### Is Claude 3.5 Haiku or DeepSeek-V3.1 better for coding?

DeepSeek-V3.1 scores higher on coding benchmarks: 40.3 versus 32.9 in the Noometry coding category.

### How many benchmarks do Claude 3.5 Haiku and DeepSeek-V3.1 share?

21 benchmarks have published results for both models. Claude 3.5 Haiku has 49 scored results on Noometry and DeepSeek-V3.1 has 27.