# Granite 4.0 H Small vs Phi 3 Mini 128k Instruct

> Granite 4.0 H Small is the stronger model overall, scoring 35.9 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/ibm-granite-h-small-vs-phi-3-mini-128k-instruct
- Last updated: 2026-10-10
- Shared benchmarks: 13

## Summary

- They share 13 benchmarks with published results for both. Granite 4.0 H Small scores higher in 8 categories and Phi 3 Mini 128k Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in instruction following, where Granite 4.0 H Small leads 68.9 to 52.1.

## Snapshot

| | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| Provider | IBM | Microsoft |
| Noometry Index | 35.9 | 29.1 |
| Rank | 213 | 303 |
| Context | — | — |
| Input $/M | — | — |
| Output $/M | — | — |
| Weights | Open | Open |

## Coding

- Granite 4.0 H Small: 35.8 (#207)
- Phi 3 Mini 128k Instruct: 28.3 (#307)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Coding | 1247 | 1040 |
| BigCodeBench Instruct | — | 29.6% |
| BigCodeBench Complete | — | 40.6% |

## Reasoning

- Granite 4.0 H Small: 23.4 (#164)
- Phi 3 Mini 128k Instruct: 18.7 (#254)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Hard Prompts | 1240 | 1028 |

## Math

- Granite 4.0 H Small: 29.3 (#219)
- Phi 3 Mini 128k Instruct: 29.2 (#220)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Math | 1251 | 1089 |
| Omni-MATH | 29.6% | — |

## Knowledge

- Granite 4.0 H Small: 31.5 (#210)
- Phi 3 Mini 128k Instruct: 26.3 (#251)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Expert | 1244 | 984 |
| MMLU-Pro | 56.9% | — |
| Vectara Hallucination Rate | 5.2% | — |
| GPQA (HELM) | 38.3% | — |

## Multilingual

- Granite 4.0 H Small: 38.7 (#223)
- Phi 3 Mini 128k Instruct: 25.4 (#280)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Non-English | 1216 | 1000 |
| LMArena Chinese | 1250 | 1016 |
| LMArena Russian | 1198 | 1004 |
| LMArena Spanish | 1258 | 1058 |
| LMArena French | — | 1038 |
| LMArena German | — | 1006 |
| LMArena Japanese | — | 899 |
| LMArena Korean | — | 856 |

## Instruction Following

- Granite 4.0 H Small: 68.9 (#178)
- Phi 3 Mini 128k Instruct: 52.1 (#289)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Instruction Following | 1221 | 1023 |
| IFEval | 89% | — |

## Long Context

- Granite 4.0 H Small: 37.6 (#209)
- Phi 3 Mini 128k Instruct: 30.3 (#284)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Longer Query | 1240 | 996 |

## Writing & Preference

- Granite 4.0 H Small: 44.0 (#222)
- Phi 3 Mini 128k Instruct: 27.3 (#296)

| Benchmark | Granite 4.0 H Small | Phi 3 Mini 128k Instruct |
|---|---|---|
| LMArena Text | 1241 | 1050 |
| LMArena Creative Writing | 1210 | 1024 |
| LMArena Multi-Turn | 1243 | 989 |
| WildBench | 73.9% | — |

## FAQ

### Is Granite 4.0 H Small better than Phi 3 Mini 128k Instruct?

Granite 4.0 H Small is the stronger model overall, scoring 35.9 to 29.1 on the Noometry Index.

### Is Granite 4.0 H Small or Phi 3 Mini 128k Instruct better for coding?

Granite 4.0 H Small scores higher on coding benchmarks: 35.8 versus 28.3 in the Noometry coding category.

### How many benchmarks do Granite 4.0 H Small and Phi 3 Mini 128k Instruct share?

13 benchmarks have published results for both models. Granite 4.0 H Small has 19 scored results on Noometry and Phi 3 Mini 128k Instruct has 19.