# Claude Haiku 4.5 vs Deepseek Coder v2

> Claude Haiku 4.5 is the stronger model overall, scoring 39.5 to 35.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-haiku-4-5-vs-deepseek-coder-v2
- Last updated: 2026-10-10
- Shared benchmarks: 17

## Summary

- They share 17 benchmarks with published results for both. Claude Haiku 4.5 scores higher in 7 categories and Deepseek Coder v2 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Claude Haiku 4.5 leads 57.9 to 38.2.
- Deepseek Coder v2 has downloadable open weights; the other is API-only.

## Snapshot

| | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| Provider | Anthropic | DeepSeek |
| Noometry Index | 39.5 | 35.9 |
| Rank | 165 | 220 |
| Context | 200K | — |
| Input $/M | $1 | — |
| Output $/M | $5 | — |
| Weights | Proprietary | Open |

## Coding

- Claude Haiku 4.5: 44.0 (#78)
- Deepseek Coder v2: 38.1 (#183)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Coding | 1453 | 1251 |
| SWE-bench Verified (bash only) | 66.6% | — |
| LMArena WebDev | 1330 | — |
| SWE-bench Multilingual | 64.7% | — |
| SciCode | 43.3% | — |
| WeirdML | 45.4% | — |
| BigCodeBench Instruct | — | 48.2% |
| BigCodeBench Complete | — | 59.7% |
| ALE-Bench | 653.48 | — |
| HumanEval+ | — | 82.3% |
| MBPP+ | — | 75.1% |

## Agentic & Tool Use

- Claude Haiku 4.5: 33.6 (#52)
- Deepseek Coder v2: —

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| Terminal-Bench | 35.5% | — |
| Berkeley Function Calling Leaderboard | 68.7% | — |
| DeepResearch Bench | 45.5% | — |
| BALROG | 31.2% | — |
| ExploitBench | 13.7% | — |
| Vending-Bench 2 | 458.89 | — |

## Reasoning

- Claude Haiku 4.5: 15.1 (#320)
- Deepseek Coder v2: 23.6 (#176)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Hard Prompts | 1420 | 1207 |
| ARC-AGI-2 | 4% | — |
| NYT Connections (extended) | 14.3% | — |
| ARC-AGI-1 | 47.7% | — |
| CritPt | 0% | — |
| Chess Puzzles | 8% | — |
| DTBench | 73.6% | — |
| LMCA | 30.9% | — |
| Epoch Capabilities Index | 142.41 | — |
| ForecastBench | 61.4 | — |
| WinoGrande | — | 83.7% |

## Math

- Claude Haiku 4.5: 44.9 (#78)
- Deepseek Coder v2: 34.9 (#190)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Math | 1396 | 1241 |
| OTIS Mock AIME 2024-2025 | 66.7% | — |
| Omni-MATH | 56.1% | — |
| MATH Level 5 | 96.4% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
| GSM8K | — | 94.5% |

## Knowledge

- Claude Haiku 4.5: 37.7 (#153)
- Deepseek Coder v2: 32.3 (#212)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Expert | 1442 | 1181 |
| GPQA Diamond | 71.2% | — |
| SimpleQA Verified | 13.2% | — |
| MMLU-Pro | 77.7% | — |
| Vectara Hallucination Rate | 9.8% | — |
| GPQA (HELM) | 60.5% | — |
| ARC (AI2) Challenge | — | 64.3% |

## Multimodal

- Claude Haiku 4.5: 26.8 (#118)
- Deepseek Coder v2: —

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| Blueprint-Bench 2 | 0% | — |
| LMArena Document | 1420 | — |

## Multilingual

- Claude Haiku 4.5: 49.9 (#129)
- Deepseek Coder v2: 36.3 (#240)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Non-English | 1377 | 1182 |
| LMArena Chinese | 1417 | 1201 |
| LMArena French | 1408 | 1185 |
| LMArena German | 1375 | 1164 |
| LMArena Japanese | 1339 | 1126 |
| LMArena Korean | 1347 | 1104 |
| LMArena Russian | 1381 | 1188 |
| LMArena Spanish | 1420 | 1153 |

## Instruction Following

- Claude Haiku 4.5: 71.4 (#149)
- Deepseek Coder v2: 61.7 (#242)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Instruction Following | 1414 | 1180 |
| IFEval | 80.1% | — |

## Long Context

- Claude Haiku 4.5: 43.6 (#92)
- Deepseek Coder v2: 37.0 (#224)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Longer Query | 1427 | 1219 |

## Writing & Preference

- Claude Haiku 4.5: 57.9 (#123)
- Deepseek Coder v2: 38.2 (#253)

| Benchmark | Claude Haiku 4.5 | Deepseek Coder v2 |
|---|---|---|
| LMArena Text | 1396 | 1191 |
| LMArena Creative Writing | 1372 | 1120 |
| LMArena Multi-Turn | 1409 | 1177 |
| WildBench | 83.9% | — |
| EQ-Bench 4 | 1064 | — |

## FAQ

### Is Claude Haiku 4.5 better than Deepseek Coder v2?

Claude Haiku 4.5 is the stronger model overall, scoring 39.5 to 35.9 on the Noometry Index.

### Is Claude Haiku 4.5 or Deepseek Coder v2 better for coding?

Claude Haiku 4.5 scores higher on coding benchmarks: 44.0 versus 38.1 in the Noometry coding category.

### How many benchmarks do Claude Haiku 4.5 and Deepseek Coder v2 share?

17 benchmarks have published results for both models. Claude Haiku 4.5 has 53 scored results on Noometry and Deepseek Coder v2 has 24.