# Claude 3 Haiku vs Gemma 3 12B

> Gemma 3 12B is the stronger model overall, scoring 32.1 to 25.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-haiku-vs-gemma-3-12b
- Last updated: 2026-10-12
- Shared benchmarks: 18

## Summary

- They share 18 benchmarks with published results for both. Claude 3 Haiku scores higher in 1 category and Gemma 3 12B in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Gemma 3 12B leads 47.5 to 29.7.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.8% for Claude 3 Haiku and 16.7% for Gemma 3 12B.
- Gemma 3 12B has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| Provider | Anthropic | Google |
| Noometry Index | 25.9 | 32.1 |
| Rank | 340 | 262 |
| Context | — | 131K |
| Input $/M | — | $0.05 |
| Output $/M | — | $0.15 |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Haiku: 26.4 (#325)
- Gemma 3 12B: 31.7 (#280)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Coding | 1199 | 1281 |
| SciCode | — | 17.4% |
| WeirdML | 9.8% | — |
| BigCodeBench Instruct | 39.4% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |

## Agentic & Tool Use

- Claude 3 Haiku: —
- Gemma 3 12B: 25.5 (#108)

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

## Reasoning

- Claude 3 Haiku: 16.3 (#307)
- Gemma 3 12B: 15.7 (#313)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Hard Prompts | 1174 | 1309 |
| DTBench | 50.1% | 48.8% |
| LMCA | 8.8% | 4.5% |
| Epoch Capabilities Index | 118.35 | 123.5 |
| Kagi LLM Benchmark | 34.2% | — |
| CritPt | — | 0% |
| Chess Puzzles | — | 0% |
| ForecastBench | 53.2 | — |
| WinoGrande | 74.2% | — |

## Math

- Claude 3 Haiku: 9.8 (#319)
- Gemma 3 12B: 22.3 (#279)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.8% | 16.7% |
| LMArena Math | 1188 | 1307 |
| MATH Level 5 | 14.9% | — |

## Knowledge

- Claude 3 Haiku: 17.3 (#285)
- Gemma 3 12B: 26.5 (#257)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| GPQA Diamond | 36.3% | 39.5% |
| LMArena Expert | 1148 | 1248 |
| Confabulations | 34.2% | — |
| Vectara Hallucination Rate | — | 4.4% |
| MMLU | 73.8% | — |

## Multimodal

- Claude 3 Haiku: 23.6 (#128)
- Gemma 3 12B: —

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Vision | 950 | — |
| MindCube | — | 46.7% |
| ScienceQA | 72% | — |

## Multilingual

- Claude 3 Haiku: 36.0 (#243)
- Gemma 3 12B: 45.7 (#165)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Non-English | 1178 | 1318 |
| LMArena German | 1174 | 1370 |
| LMArena Russian | 1204 | 1335 |
| LMArena Chinese | 1155 | — |
| LMArena French | 1195 | — |
| LMArena Japanese | 1102 | — |
| LMArena Korean | 1109 | — |
| LMArena Spanish | 1166 | — |

## Instruction Following

- Claude 3 Haiku: 61.3 (#247)
- Gemma 3 12B: 68.6 (#186)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Instruction Following | 1173 | 1299 |

## Long Context

- Claude 3 Haiku: 36.1 (#237)
- Gemma 3 12B: 40.0 (#162)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Longer Query | 1190 | 1317 |

## Writing & Preference

- Claude 3 Haiku: 29.7 (#291)
- Gemma 3 12B: 47.5 (#209)

| Benchmark | Claude 3 Haiku | Gemma 3 12B |
|---|---|---|
| LMArena Text | 1195 | 1334 |
| LMArena Creative Writing | 1157 | 1331 |
| EQ-Bench Creative Writing | 717 | 1126 |
| LMArena Multi-Turn | 1190 | 1334 |

## FAQ

### Is Claude 3 Haiku better than Gemma 3 12B?

Gemma 3 12B is the stronger model overall, scoring 32.1 to 25.9 on the Noometry Index.

### Is Claude 3 Haiku or Gemma 3 12B better for coding?

Gemma 3 12B scores higher on coding benchmarks: 31.7 versus 26.4 in the Noometry coding category.

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

18 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Gemma 3 12B has 24.