Model comparison
Grok-3 mini vs Granite 4.0 H Small
Grok-3 mini is the stronger model overall, scoring 41.0 to 35.9 on the Noometry Index.
Last verified . 18 shared benchmarks.
Summary
- They share 18 benchmarks with published results for both. Grok-3 mini scores higher in 7 categories and Granite 4.0 H Small in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok-3 mini leads 46.3 to 31.5.
- The biggest single-benchmark swing is GPQA (HELM): 67.5% for Grok-3 mini and 38.3% for Granite 4.0 H Small.
- Granite 4.0 H Small has downloadable open weights; the other is API-only.
Side by side
| Grok-3 mini | Granite 4.0 H Small | |
|---|---|---|
| Provider | xAI | IBM |
| Noometry Index | 41.0 | 35.9 |
| Released | 2025-04-09 | — |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 35 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok-3 mini leads
Grok-3 mini: 40.4 (#126), Granite 4.0 H Small: 35.8 (#207)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| LMArena Coding | 1380 | 1247 |
| Aider Polyglot | 49.3% | — |
| WeirdML | 42.6% | — |
Reasoning Granite 4.0 H Small leads
Grok-3 mini: 13.8 (#323), Granite 4.0 H Small: 23.4 (#164)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| LMArena Hard Prompts | 1374 | 1240 |
| ARC-AGI-2 | 0.4% | — |
| Kagi LLM Benchmark | 61.3% | — |
| ARC-AGI-1 | 16.5% | — |
| Epoch Capabilities Index | 140.35 | — |
Math Grok-3 mini leads
Grok-3 mini: 40.3 (#78), Granite 4.0 H Small: 29.3 (#219)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| Omni-MATH | 31.8% | 29.6% |
| LMArena Math | 1386 | 1251 |
| OTIS Mock AIME 2024-2025 | 77.8% | — |
| MATH Level 5 | 90.9% | — |
| FrontierMath (Feb 2025 set) | 5.9% | — |
Knowledge Grok-3 mini leads
Grok-3 mini: 46.3 (#75), Granite 4.0 H Small: 31.5 (#210)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| MMLU-Pro | 79.9% | 56.9% |
| GPQA (HELM) | 67.5% | 38.3% |
| LMArena Expert | 1395 | 1244 |
| GPQA Diamond | 76.3% | — |
| Confabulations | 10.8% | — |
| Vectara Hallucination Rate | — | 5.2% |
Multilingual Grok-3 mini leads
Grok-3 mini: 48.3 (#141), Granite 4.0 H Small: 38.7 (#223)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| LMArena Non-English | 1352 | 1216 |
| LMArena Chinese | 1387 | 1250 |
| LMArena Russian | 1352 | 1198 |
| LMArena Spanish | 1379 | 1258 |
| LMArena French | 1357 | — |
| LMArena German | 1350 | — |
| LMArena Japanese | 1343 | — |
| LMArena Korean | 1336 | — |
Instruction Following Grok-3 mini leads
Grok-3 mini: 78.7 (#8), Granite 4.0 H Small: 68.9 (#178)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| IFEval | 95.1% | 89% |
| LMArena Instruction Following | 1356 | 1221 |
Long Context Grok-3 mini leads
Grok-3 mini: 41.0 (#142), Granite 4.0 H Small: 37.6 (#209)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| LMArena Longer Query | 1372 | 1240 |
| Fiction.LiveBench | 66.7% | — |
Writing & Preference Grok-3 mini leads
Grok-3 mini: 52.7 (#163), Granite 4.0 H Small: 44.0 (#222)
| Benchmark | Grok-3 mini | Granite 4.0 H Small |
|---|---|---|
| LMArena Text | 1370 | 1241 |
| LMArena Creative Writing | 1342 | 1210 |
| WildBench | 65.1% | 73.9% |
| LMArena Multi-Turn | 1355 | 1243 |
| Short-Story Creative Writing | 73.5% | — |
Frequently asked questions
Is Grok-3 mini better than Granite 4.0 H Small?
Grok-3 mini is the stronger model overall, scoring 41.0 to 35.9 on the Noometry Index.
Is Grok-3 mini or Granite 4.0 H Small better for coding?
Grok-3 mini scores higher on coding benchmarks: 40.4 versus 35.8 in the Noometry coding category.
How many benchmarks do Grok-3 mini and Granite 4.0 H Small share?
18 benchmarks have published results for both models. Grok-3 mini has 35 scored results on Noometry and Granite 4.0 H Small has 19.