# Gemma 3 27B vs o4-mini

> o4-mini is the stronger model overall, scoring 41.6 to 30.5 on the Noometry Index. Gemma 3 27B costs 19× less per token, which makes it the better buy when o4-mini's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gemma-3-27b-vs-o4-mini
- Last updated: 2026-10-10
- Shared benchmarks: 34

## Summary

- They share 34 benchmarks with published results for both. Gemma 3 27B scores higher in 0 categories and o4-mini in 10 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in coding, where o4-mini leads 41.6 to 22.1.
- The biggest single-benchmark swing is Aider Polyglot: 4.9% for Gemma 3 27B and 72% for o4-mini.
- Gemma 3 27B is cheaper at $0.08 / $0.16 per million input/output tokens, against $1.10 / $4.40 for o4-mini.
- o4-mini accepts more context: 200K tokens versus 131K.
- Gemma 3 27B has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 3 27B | o4-mini |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 30.5 | 41.6 |
| Rank | 277 | 119 |
| Context | 131K | 200K |
| Input $/M | $0.08 | $1.10 |
| Output $/M | $0.16 | $4.40 |
| Weights | Open | Proprietary |

## Coding

- Gemma 3 27B: 22.1 (#329)
- o4-mini: 41.6 (#110)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| Aider Polyglot | 4.9% | 72% |
| LMArena Coding | 1322 | 1369 |
| SWE-bench Verified (bash only) | — | 45% |
| SciCode | 21.2% | — |
| GSO | — | 3.6% |
| WeirdML | — | 52.6% |
| LiveBench Coding | 39.9% | — |
| CadEval | — | 62% |
| ALE-Bench | — | 826.17 |
| AlgoTune | — | 1.72 |

## Agentic & Tool Use

- Gemma 3 27B: 24.9 (#105)
- o4-mini: 32.3 (#57)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| Berkeley Function Calling Leaderboard | 29.5% | 53.2% |
| GDPval | — | 25.3% |
| METR Time Horizons | — | 63.9% |

## Reasoning

- Gemma 3 27B: 16.3 (#297)
- o4-mini: 24.6 (#143)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| Kagi LLM Benchmark | 40.4% | 67.6% |
| CritPt | 0% | 0.6% |
| Chess Puzzles | 0% | 26% |
| LMArena Hard Prompts | 1340 | 1351 |
| DTBench | 52.5% | 77.6% |
| LMCA | 12.3% | 26.5% |
| Epoch Capabilities Index | 130.04 | 145.63 |
| ARC-AGI-2 | — | 6.1% |
| SimpleBench | — | 38.7% |
| ARC-AGI-1 | — | 58.7% |
| EnigmaEval | — | 9.2% |
| LiveBench Reasoning | 43.8% | — |
| Mystery Game Puzzles | — | 5% |
| LiveBench Data Analysis | 51.5% | — |
| ForecastBench | — | 61.8 |
| LiveBench | 50% | — |

## Math

- Gemma 3 27B: 24.4 (#261)
- o4-mini: 40.2 (#79)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 22.5% | 81.7% |
| LMArena Math | 1312 | 1390 |
| MATH Level 5 | 74% | 97.8% |
| FrontierMath (Tiers 1-3) | — | 36.1% |
| FrontierMath Tier 4 | — | 4.9% |
| Omni-MATH | — | 72% |
| LiveBench Math | 55.4% | — |
| FrontierMath (Feb 2025 set) | — | 24.8% |
| FrontierMath Tier 4 (v1) | — | 6.3% |

## Knowledge

- Gemma 3 27B: 25.4 (#256)
- o4-mini: 43.5 (#83)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| GPQA Diamond | 47.7% | 79.6% |
| Confabulations | 40.3% | 15.8% |
| Vectara Hallucination Rate | 7.4% | 18.6% |
| LMArena Expert | 1303 | 1344 |
| Humanity's Last Exam | — | 18.1% |
| SimpleQA Verified | — | 19.6% |
| MMLU-Pro | — | 82% |
| GPQA (HELM) | — | 73.5% |

## Multimodal

- Gemma 3 27B: 32.0 (#94)
- o4-mini: 39.5 (#46)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| LMArena Vision | 1165 | 1194 |
| GeoBench | 52% | 64% |
| VPCT | — | 57.5% |

## Multilingual

- Gemma 3 27B: 47.0 (#150)
- o4-mini: 47.2 (#149)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| LMArena Non-English | 1334 | 1336 |
| LMArena Chinese | 1347 | 1355 |
| LMArena French | 1368 | 1362 |
| LMArena German | 1362 | 1336 |
| LMArena Japanese | 1287 | 1308 |
| LMArena Korean | 1308 | 1313 |
| LMArena Russian | 1349 | 1334 |
| LMArena Spanish | 1348 | 1346 |

## Instruction Following

- Gemma 3 27B: 70.8 (#155)
- o4-mini: 75.4 (#63)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| LMArena Instruction Following | 1321 | 1321 |
| LiveBench Instruction Following | 74.9% | — |
| IFEval | — | 92.8% |

## Long Context

- Gemma 3 27B: 27.6 (#288)
- o4-mini: 45.5 (#31)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| Fiction.LiveBench | 33.3% | 77.8% |
| LMArena Longer Query | 1333 | 1315 |

## Writing & Preference

- Gemma 3 27B: 52.7 (#161)
- o4-mini: 54.2 (#148)

| Benchmark | Gemma 3 27B | o4-mini |
|---|---|---|
| LMArena Text | 1358 | 1353 |
| LMArena Creative Writing | 1346 | 1294 |
| Short-Story Creative Writing | 79.9% | 75% |
| LMArena Multi-Turn | 1344 | 1350 |
| EQ-Bench Creative Writing | 1266 | — |
| WildBench | — | 85.4% |
| LiveBench Language | 34.6% | — |

## FAQ

### Is Gemma 3 27B better than o4-mini?

o4-mini is the stronger model overall, scoring 41.6 to 30.5 on the Noometry Index. Gemma 3 27B costs 19× less per token, which makes it the better buy when o4-mini's lead doesn't matter for your workload.

### Which is cheaper, Gemma 3 27B or o4-mini?

Gemma 3 27B is cheaper. It lists at $0.08 per million input tokens and $0.16 per million output tokens; o4-mini lists at $1.10 and $4.40.

### Is Gemma 3 27B or o4-mini better for coding?

o4-mini scores higher on coding benchmarks: 41.6 versus 22.1 in the Noometry coding category.

### Which has the bigger context window?

o4-mini does, with 200K tokens against 131K.

### How many benchmarks do Gemma 3 27B and o4-mini share?

34 benchmarks have published results for both models. Gemma 3 27B has 43 scored results on Noometry and o4-mini has 60.