# Claude 3 Haiku vs Granite 3.0 2b Instruct

> Granite 3.0 2b Instruct is the stronger model overall, scoring 30.8 to 25.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-haiku-vs-granite-3-0-2b-instruct
- Last updated: 2026-10-12
- Shared benchmarks: 13

## Summary

- They share 13 benchmarks with published results for both. Claude 3 Haiku scores higher in 4 categories and Granite 3.0 2b Instruct in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where Granite 3.0 2b Instruct leads 32.2 to 9.8.
- The biggest single-benchmark swing is BigCodeBench Instruct: 39.4% for Claude 3 Haiku and 20.5% for Granite 3.0 2b Instruct.
- Granite 3.0 2b Instruct has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| Provider | Anthropic | IBM |
| Noometry Index | 25.9 | 30.8 |
| Rank | 340 | 286 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Haiku: 26.4 (#325)
- Granite 3.0 2b Instruct: 28.3 (#316)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| BigCodeBench Instruct | 39.4% | 20.5% |
| LMArena Coding | 1199 | 1090 |
| WeirdML | 9.8% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |

## Reasoning

- Claude 3 Haiku: 16.3 (#307)
- Granite 3.0 2b Instruct: 20.5 (#235)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Hard Prompts | 1174 | 1073 |
| Kagi LLM Benchmark | 34.2% | — |
| DTBench | 50.1% | — |
| LMCA | 8.8% | — |
| Epoch Capabilities Index | 118.35 | — |
| ForecastBench | 53.2 | — |
| WinoGrande | 74.2% | — |

## Math

- Claude 3 Haiku: 9.8 (#319)
- Granite 3.0 2b Instruct: 32.2 (#217)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Math | 1188 | 1117 |
| OTIS Mock AIME 2024-2025 | 1.8% | — |
| MATH Level 5 | 14.9% | — |

## Knowledge

- Claude 3 Haiku: 17.3 (#285)
- Granite 3.0 2b Instruct: 29.0 (#241)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Expert | 1148 | 1064 |
| GPQA Diamond | 36.3% | — |
| Confabulations | 34.2% | — |
| MMLU | 73.8% | — |

## Multimodal

- Claude 3 Haiku: 23.6 (#128)
- Granite 3.0 2b Instruct: —

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Vision | 950 | — |
| ScienceQA | 72% | — |

## Multilingual

- Claude 3 Haiku: 36.0 (#243)
- Granite 3.0 2b Instruct: 27.0 (#278)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Non-English | 1178 | 1033 |
| LMArena Chinese | 1155 | 1070 |
| LMArena Russian | 1204 | 1045 |
| LMArena French | 1195 | — |
| LMArena German | 1174 | — |
| LMArena Japanese | 1102 | — |
| LMArena Korean | 1109 | — |
| LMArena Spanish | 1166 | — |

## Instruction Following

- Claude 3 Haiku: 61.3 (#247)
- Granite 3.0 2b Instruct: 53.9 (#284)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Instruction Following | 1173 | 1056 |

## Long Context

- Claude 3 Haiku: 36.1 (#237)
- Granite 3.0 2b Instruct: 32.5 (#268)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Longer Query | 1190 | 1070 |

## Writing & Preference

- Claude 3 Haiku: 29.7 (#291)
- Granite 3.0 2b Instruct: 29.6 (#292)

| Benchmark | Claude 3 Haiku | Granite 3.0 2b Instruct |
|---|---|---|
| LMArena Text | 1195 | 1080 |
| LMArena Creative Writing | 1157 | 1046 |
| LMArena Multi-Turn | 1190 | 1053 |
| EQ-Bench Creative Writing | 717 | — |

## FAQ

### Is Claude 3 Haiku better than Granite 3.0 2b Instruct?

Granite 3.0 2b Instruct is the stronger model overall, scoring 30.8 to 25.9 on the Noometry Index.

### Is Claude 3 Haiku or Granite 3.0 2b Instruct better for coding?

Granite 3.0 2b Instruct scores higher on coding benchmarks: 28.3 versus 26.4 in the Noometry coding category.

### How many benchmarks do Claude 3 Haiku and Granite 3.0 2b Instruct share?

13 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Granite 3.0 2b Instruct has 13.