# Gemma 3 12B vs Grok-3 mini

> Grok-3 mini is the stronger model overall, scoring 41.2 to 32.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemma-3-12b-vs-grok-3-mini
- Last updated: 2026-10-10
- Shared benchmarks: 15

## Summary

- They share 15 benchmarks with published results for both. Gemma 3 12B scores higher in 1 category and Grok-3 mini in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok-3 mini leads 46.4 to 26.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 16.7% for Gemma 3 12B and 77.8% for Grok-3 mini.
- Gemma 3 12B has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| Provider | Google | xAI |
| Noometry Index | 32.1 | 41.2 |
| Rank | 262 | 141 |
| Context | 131K | — |
| Input $/M | $0.05 | — |
| Output $/M | $0.15 | — |
| Weights | Open | Proprietary |

## Coding

- Gemma 3 12B: 31.7 (#280)
- Grok-3 mini: 40.8 (#131)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| LMArena Coding | 1281 | 1379 |
| Aider Polyglot | — | 49.3% |
| SciCode | 17.4% | — |
| WeirdML | — | 42.6% |

## Agentic & Tool Use

- Gemma 3 12B: 25.5 (#108)
- Grok-3 mini: —

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 30.4% | — |

## Reasoning

- Gemma 3 12B: 15.7 (#313)
- Grok-3 mini: 13.6 (#334)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| LMArena Hard Prompts | 1309 | 1375 |
| Epoch Capabilities Index | 123.5 | 140.35 |
| ARC-AGI-2 | — | 0.4% |
| Kagi LLM Benchmark | — | 61.3% |
| ARC-AGI-1 | — | 16.5% |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| DTBench | 48.8% | — |
| LMCA | 4.5% | — |

## Math

- Gemma 3 12B: 22.3 (#279)
- Grok-3 mini: 42.1 (#85)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 16.7% | 77.8% |
| LMArena Math | 1307 | 1386 |
| Omni-MATH | — | 31.8% |
| MATH Level 5 | — | 90.9% |
| FrontierMath (Feb 2025 set) | — | 5.9% |

## Knowledge

- Gemma 3 12B: 26.5 (#257)
- Grok-3 mini: 46.4 (#81)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| GPQA Diamond | 39.5% | 76.3% |
| LMArena Expert | 1248 | 1395 |
| MMLU-Pro | — | 79.9% |
| Confabulations | — | 10.8% |
| Vectara Hallucination Rate | 4.4% | — |
| GPQA (HELM) | — | 67.5% |

## Multimodal

- Gemma 3 12B: —
- Grok-3 mini: —

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| MindCube | 46.7% | — |

## Multilingual

- Gemma 3 12B: 45.7 (#165)
- Grok-3 mini: 48.1 (#145)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| LMArena Non-English | 1318 | 1352 |
| LMArena German | 1370 | 1349 |
| LMArena Russian | 1335 | 1353 |
| LMArena Chinese | — | 1387 |
| LMArena French | — | 1357 |
| LMArena Japanese | — | 1342 |
| LMArena Korean | — | 1335 |
| LMArena Spanish | — | 1381 |

## Instruction Following

- Gemma 3 12B: 68.6 (#186)
- Grok-3 mini: 78.5 (#9)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| LMArena Instruction Following | 1299 | 1357 |
| IFEval | — | 95.1% |

## Long Context

- Gemma 3 12B: 40.0 (#162)
- Grok-3 mini: 41.0 (#147)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| LMArena Longer Query | 1317 | 1372 |
| Fiction.LiveBench | — | 66.7% |

## Writing & Preference

- Gemma 3 12B: 47.5 (#209)
- Grok-3 mini: 52.5 (#169)

| Benchmark | Gemma 3 12B | Grok-3 mini |
|---|---|---|
| LMArena Text | 1334 | 1370 |
| LMArena Creative Writing | 1331 | 1342 |
| LMArena Multi-Turn | 1334 | 1355 |
| Short-Story Creative Writing | — | 73.5% |
| EQ-Bench Creative Writing | 1126 | — |
| WildBench | — | 65.1% |

## FAQ

### Is Gemma 3 12B better than Grok-3 mini?

Grok-3 mini is the stronger model overall, scoring 41.2 to 32.1 on the Noometry Index.

### Is Gemma 3 12B or Grok-3 mini better for coding?

Grok-3 mini scores higher on coding benchmarks: 40.8 versus 31.7 in the Noometry coding category.

### How many benchmarks do Gemma 3 12B and Grok-3 mini share?

15 benchmarks have published results for both models. Gemma 3 12B has 24 scored results on Noometry and Grok-3 mini has 35.