# Granite 4.0 H Small vs Mistral Large 3

> Mistral Large 3 is the stronger model overall, scoring 38.6 to 35.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/ibm-granite-h-small-vs-mistral-large-3
- Last updated: 2026-10-10
- Shared benchmarks: 14

## Summary

- They share 14 benchmarks with published results for both. Granite 4.0 H Small scores higher in 2 categories and Mistral Large 3 in 6 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Mistral Large 3 leads 60.2 to 44.0.
- The biggest single-benchmark swing is Vectara Hallucination Rate: 5.2% for Granite 4.0 H Small and 14.5% for Mistral Large 3.

## Snapshot

| | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| Provider | IBM | Mistral AI |
| Noometry Index | 35.9 | 38.6 |
| Rank | 213 | 172 |
| Context | — | 262K |
| Input $/M | — | $0.25 |
| Output $/M | — | $0.75 |
| Weights | Open | Open |

## Coding

- Granite 4.0 H Small: 35.8 (#207)
- Mistral Large 3: 34.1 (#230)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Coding | 1247 | 1447 |
| LMArena WebDev | — | 1230 |

## Reasoning

- Granite 4.0 H Small: 23.4 (#164)
- Mistral Large 3: 14.6 (#314)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Hard Prompts | 1240 | 1429 |
| Kagi LLM Benchmark | — | 50.9% |
| NYT Connections (extended) | — | 7.5% |
| Thematic Generalization | — | 23% |

## Math

- Granite 4.0 H Small: 29.3 (#219)
- Mistral Large 3: 36.0 (#131)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Math | 1251 | 1415 |
| Omni-MATH | 29.6% | — |

## Knowledge

- Granite 4.0 H Small: 31.5 (#210)
- Mistral Large 3: 35.5 (#173)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| Vectara Hallucination Rate | 5.2% | 14.5% |
| LMArena Expert | 1244 | 1421 |
| MMLU-Pro | 56.9% | — |
| GPQA (HELM) | 38.3% | — |

## Multimodal

- Granite 4.0 H Small: —
- Mistral Large 3: 37.6 (#65)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Vision | — | 1222 |

## Multilingual

- Granite 4.0 H Small: 38.7 (#223)
- Mistral Large 3: 52.6 (#79)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Non-English | 1216 | 1412 |
| LMArena Chinese | 1250 | 1448 |
| LMArena Russian | 1198 | 1412 |
| LMArena Spanish | 1258 | 1437 |
| LMArena French | — | 1456 |
| LMArena German | — | 1438 |
| LMArena Japanese | — | 1399 |
| LMArena Korean | — | 1384 |

## Instruction Following

- Granite 4.0 H Small: 68.9 (#178)
- Mistral Large 3: 74.1 (#104)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Instruction Following | 1221 | 1402 |
| IFEval | 89% | — |

## Long Context

- Granite 4.0 H Small: 37.6 (#209)
- Mistral Large 3: 43.1 (#102)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Longer Query | 1240 | 1413 |

## Writing & Preference

- Granite 4.0 H Small: 44.0 (#222)
- Mistral Large 3: 60.2 (#100)

| Benchmark | Granite 4.0 H Small | Mistral Large 3 |
|---|---|---|
| LMArena Text | 1241 | 1427 |
| LMArena Creative Writing | 1210 | 1385 |
| LMArena Multi-Turn | 1243 | 1429 |
| EQ-Bench Creative Writing | — | 1412 |
| WildBench | 73.9% | — |

## FAQ

### Is Granite 4.0 H Small better than Mistral Large 3?

Mistral Large 3 is the stronger model overall, scoring 38.6 to 35.9 on the Noometry Index.

### Is Granite 4.0 H Small or Mistral Large 3 better for coding?

Granite 4.0 H Small scores higher on coding benchmarks: 35.8 versus 34.1 in the Noometry coding category.

### How many benchmarks do Granite 4.0 H Small and Mistral Large 3 share?

14 benchmarks have published results for both models. Granite 4.0 H Small has 19 scored results on Noometry and Mistral Large 3 has 24.