# Gemma 3 27B vs Grok 3

> Grok 3 is the stronger model overall, scoring 39.6 to 30.5 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemma-3-27b-vs-grok-3
- Last updated: 2026-10-10
- Shared benchmarks: 28

## Summary

- They share 28 benchmarks with published results for both. Gemma 3 27B scores higher in 1 category and Grok 3 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 3 leads 46.0 to 25.4.
- The biggest single-benchmark swing is Aider Polyglot: 4.9% for Gemma 3 27B and 53.3% for Grok 3.
- Gemma 3 27B has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 3 27B | Grok 3 |
|---|---|---|
| Provider | Google | xAI |
| Noometry Index | 30.5 | 39.6 |
| Rank | 277 | 150 |
| Context | 131K | — |
| Input $/M | $0.08 | — |
| Output $/M | $0.16 | — |
| Weights | Open | Proprietary |

## Coding

- Gemma 3 27B: 22.1 (#329)
- Grok 3: 41.5 (#113)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| Aider Polyglot | 4.9% | 53.3% |
| LMArena Coding | 1322 | 1432 |
| SciCode | 21.2% | — |
| WeirdML | — | 37.2% |
| LiveBench Coding | 39.9% | — |

## Agentic & Tool Use

- Gemma 3 27B: 24.9 (#105)
- Grok 3: 30.2 (#69)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | — |
| BALROG | — | 29.5% |

## Reasoning

- Gemma 3 27B: 16.3 (#297)
- Grok 3: 13.8 (#322)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 61.3% |
| LMArena Hard Prompts | 1340 | 1433 |
| Epoch Capabilities Index | 130.04 | 138.34 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 36.1% |
| ARC-AGI-1 | — | 5.5% |
| CritPt | 0% | — |
| Chess Puzzles | 0% | — |
| LiveBench Reasoning | 43.8% | — |
| DTBench | 52.5% | — |
| LiveBench Data Analysis | 51.5% | — |
| LMCA | 12.3% | — |
| LiveBench | 50% | — |

## Math

- Gemma 3 27B: 24.4 (#261)
- Grok 3: 36.2 (#122)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 55.6% |
| LMArena Math | 1312 | 1390 |
| MATH Level 5 | 74% | 88.7% |
| Omni-MATH | — | 46.4% |
| LiveBench Math | 55.4% | — |
| FrontierMath (Feb 2025 set) | — | 3.8% |
| FrontierMath Tier 4 (v1) | — | 0% |

## Knowledge

- Gemma 3 27B: 25.4 (#256)
- Grok 3: 46.0 (#76)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| GPQA Diamond | 47.7% | 75.8% |
| Confabulations | 40.3% | 14.2% |
| Vectara Hallucination Rate | 7.4% | 5.8% |
| LMArena Expert | 1303 | 1420 |
| MMLU-Pro | — | 78.8% |
| GPQA (HELM) | — | 65% |

## Multimodal

- Gemma 3 27B: 32.0 (#94)
- Grok 3: —

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| LMArena Vision | 1165 | — |
| GeoBench | 52% | — |

## Multilingual

- Gemma 3 27B: 47.0 (#150)
- Grok 3: 52.5 (#81)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| LMArena Non-English | 1334 | 1411 |
| LMArena Chinese | 1347 | 1448 |
| LMArena French | 1368 | 1460 |
| LMArena German | 1362 | 1431 |
| LMArena Japanese | 1287 | 1389 |
| LMArena Korean | 1308 | 1374 |
| LMArena Russian | 1349 | 1416 |
| LMArena Spanish | 1348 | 1417 |

## Instruction Following

- Gemma 3 27B: 70.8 (#155)
- Grok 3: 75.2 (#69)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| LMArena Instruction Following | 1321 | 1409 |
| LiveBench Instruction Following | 74.9% | — |
| IFEval | — | 88.4% |

## Long Context

- Gemma 3 27B: 27.6 (#288)
- Grok 3: 38.7 (#187)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| Fiction.LiveBench | 33.3% | 58.3% |
| LMArena Longer Query | 1333 | 1439 |

## Writing & Preference

- Gemma 3 27B: 52.7 (#161)
- Grok 3: 56.0 (#136)

| Benchmark | Gemma 3 27B | Grok 3 |
|---|---|---|
| LMArena Text | 1358 | 1426 |
| LMArena Creative Writing | 1346 | 1414 |
| Short-Story Creative Writing | 79.9% | 76.4% |
| EQ-Bench Creative Writing | 1266 | 1186 |
| LMArena Multi-Turn | 1344 | 1424 |
| WildBench | — | 84.9% |
| LiveBench Language | 34.6% | — |

## FAQ

### Is Gemma 3 27B better than Grok 3?

Grok 3 is the stronger model overall, scoring 39.6 to 30.5 on the Noometry Index.

### Is Gemma 3 27B or Grok 3 better for coding?

Grok 3 scores higher on coding benchmarks: 41.5 versus 22.1 in the Noometry coding category.

### How many benchmarks do Gemma 3 27B and Grok 3 share?

28 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and Grok 3 has 40.