# Gemma 2B vs Gemma 3 12B

> Gemma 3 12B is the stronger model overall, scoring 31.8 to 29.1 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemma-2b-vs-gemma-3-12b
- Last updated: 2026-10-10
- Shared benchmarks: 11

## Summary

- They share 11 benchmarks with published results for both. Gemma 2B scores higher in 2 categories and Gemma 3 12B in 5 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 12B leads 47.7 to 24.1.

## Snapshot

| | Gemma 2B | Gemma 3 12B |
|---|---|---|
| Provider | Google | Google |
| Noometry Index | 29.1 | 31.8 |
| Rank | 305 | 255 |
| Context | — | 131K |
| Input $/M | — | $0.05 |
| Output $/M | — | $0.15 |
| Weights | Open | Open |

## Coding

- Gemma 2B: 28.9 (#300)
- Gemma 3 12B: 31.2 (#276)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Coding | 1010 | 1281 |
| SciCode | — | 17.4% |
| HumanEval+ | 20.7% | — |
| MBPP+ | 34.1% | — |

## Agentic & Tool Use

- Gemma 2B: —
- Gemma 3 12B: 25.3 (#101)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 30.4% |

## Reasoning

- Gemma 2B: 17.9 (#275)
- Gemma 3 12B: 15.2 (#307)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Hard Prompts | 989 | 1309 |
| Epoch Capabilities Index | 93.57 | 123.49 |
| CritPt | — | 0% |
| Chess Puzzles | — | 0% |
| DTBench | — | 48.8% |
| LMCA | — | 4.5% |
| BIG-Bench Hard | 35.2% | — |
| HellaSwag | 71.4% | — |
| PIQA | 77.3% | — |
| WinoGrande | 65.4% | — |

## Math

- Gemma 2B: 27.6 (#237)
- Gemma 3 12B: 21.0 (#273)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Math | 1009 | 1307 |
| OTIS Mock AIME 2024-2025 | — | 16.7% |
| GSM8K | 17.7% | — |

## Knowledge

- Gemma 2B: —
- Gemma 3 12B: 26.5 (#250)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| GPQA Diamond | — | 39.5% |
| Vectara Hallucination Rate | — | 4.4% |
| LMArena Expert | — | 1248 |
| ARC (AI2) Challenge | 42.1% | — |
| BoolQ | 69.4% | — |
| MMLU | 42.3% | — |
| TriviaQA | 53.2% | — |

## Multimodal

- Gemma 2B: —
- Gemma 3 12B: —

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| MindCube | — | 46.7% |

## Multilingual

- Gemma 2B: 23.1 (#289)
- Gemma 3 12B: 45.9 (#160)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Non-English | 958 | 1318 |
| LMArena Russian | 937 | 1335 |
| LMArena Chinese | 986 | — |
| LMArena German | — | 1370 |

## Instruction Following

- Gemma 2B: 48.8 (#297)
- Gemma 3 12B: 68.8 (#179)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Instruction Following | 971 | 1299 |

## Long Context

- Gemma 2B: 29.9 (#286)
- Gemma 3 12B: 40.0 (#157)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Longer Query | 981 | 1317 |

## Writing & Preference

- Gemma 2B: 24.1 (#303)
- Gemma 3 12B: 47.7 (#205)

| Benchmark | Gemma 2B | Gemma 3 12B |
|---|---|---|
| LMArena Text | 1002 | 1334 |
| LMArena Creative Writing | 986 | 1331 |
| LMArena Multi-Turn | 945 | 1334 |
| EQ-Bench Creative Writing | — | 1126 |

## FAQ

### Is Gemma 2B better than Gemma 3 12B?

Gemma 3 12B is the stronger model overall, scoring 31.8 to 29.1 on the Noometry Index.

### Is Gemma 2B or Gemma 3 12B better for coding?

Gemma 3 12B scores higher on coding benchmarks: 31.2 versus 28.9 in the Noometry coding category.

### How many benchmarks do Gemma 2B and Gemma 3 12B share?

11 benchmarks have published results for both models. Gemma 2B has 23 scored results on Noometry and Gemma 3 12B has 24.