# Claude 3 Haiku vs DeepSeek V4.1 Flash

> DeepSeek V4.1 Flash is the stronger model overall, scoring 50.3 to 25.7 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-haiku-vs-deepseek-v4-1-flash
- Last updated: 2026-10-10
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Claude 3 Haiku scores higher in 0 categories and DeepSeek V4.1 Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where DeepSeek V4.1 Flash leads 48.4 to 9.2.
- The biggest single-benchmark swing is DTBench: 50.1% for Claude 3 Haiku and 89.9% for DeepSeek V4.1 Flash.
- DeepSeek V4.1 Flash has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 25.7 | 50.3 |
| Rank | 333 | 40 |
| Context | — | 1M |
| Input $/M | — | $0.15 |
| Output $/M | — | $0.60 |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Haiku: 26.1 (#320)
- DeepSeek V4.1 Flash: 52.8 (#29)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Coding | 1199 | 1511 |
| LMArena WebDev | — | 1620 |
| SciCode | — | 51.9% |
| WeirdML | 9.8% | — |
| BigCodeBench Instruct | 39.4% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| ALE-Bench | — | 1,092 |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |

## Reasoning

- Claude 3 Haiku: 15.7 (#303)
- DeepSeek V4.1 Flash: 46.2 (#39)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Hard Prompts | 1174 | 1485 |
| DTBench | 50.1% | 89.9% |
| LMCA | 8.8% | 47% |
| Epoch Capabilities Index | 118.3 | 155.01 |
| Kagi LLM Benchmark | 34.2% | — |
| NYT Connections (extended) | — | 89.6% |
| CritPt | — | 14.3% |
| Surface Evolver Bench | — | 46.3% |
| ForecastBench | 53.2 | — |
| WinoGrande | 74.2% | — |

## Math

- Claude 3 Haiku: 9.2 (#313)
- DeepSeek V4.1 Flash: 48.4 (#52)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Math | 1188 | 1493 |
| OTIS Mock AIME 2024-2025 | 1.8% | — |
| ProofBench | — | 54% |
| MATH Level 5 | 14.9% | — |

## Knowledge

- Claude 3 Haiku: 17.3 (#280)
- DeepSeek V4.1 Flash: 41.6 (#97)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Expert | 1148 | 1505 |
| GPQA Diamond | 36.3% | — |
| Confabulations | 34.2% | — |
| MMLU | 73.8% | — |

## Multimodal

- Claude 3 Haiku: 23.1 (#123)
- DeepSeek V4.1 Flash: —

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Vision | 950 | — |
| ScienceQA | 72% | — |

## Multilingual

- Claude 3 Haiku: 36.1 (#238)
- DeepSeek V4.1 Flash: 55.5 (#29)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Non-English | 1178 | 1452 |
| LMArena Chinese | 1155 | 1506 |
| LMArena French | 1194 | 1448 |
| LMArena Russian | 1204 | 1473 |
| LMArena Spanish | 1165 | 1499 |
| LMArena German | 1174 | — |
| LMArena Japanese | 1102 | — |
| LMArena Korean | 1108 | — |

## Instruction Following

- Claude 3 Haiku: 61.5 (#242)
- DeepSeek V4.1 Flash: 77.6 (#24)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Instruction Following | 1174 | 1475 |

## Long Context

- Claude 3 Haiku: 36.1 (#232)
- DeepSeek V4.1 Flash: 45.2 (#46)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Longer Query | 1191 | 1476 |

## Writing & Preference

- Claude 3 Haiku: 29.8 (#286)
- DeepSeek V4.1 Flash: 65.8 (#43)

| Benchmark | Claude 3 Haiku | DeepSeek V4.1 Flash |
|---|---|---|
| LMArena Text | 1195 | 1465 |
| LMArena Creative Writing | 1157 | 1436 |
| EQ-Bench Creative Writing | 717 | 1540 |
| LMArena Multi-Turn | 1190 | 1460 |

## FAQ

### Is Claude 3 Haiku better than DeepSeek V4.1 Flash?

DeepSeek V4.1 Flash is the stronger model overall, scoring 50.3 to 25.7 on the Noometry Index.

### Is Claude 3 Haiku or DeepSeek V4.1 Flash better for coding?

DeepSeek V4.1 Flash scores higher on coding benchmarks: 52.8 versus 26.1 in the Noometry coding category.

### How many benchmarks do Claude 3 Haiku and DeepSeek V4.1 Flash share?

18 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and DeepSeek V4.1 Flash has 25.