# Claude 3 Haiku vs Gemma 4 31B IT

> Gemma 4 31B IT is the stronger model overall, scoring 43.5 to 25.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/claude-3-haiku-vs-gemma-4-31b-it
- Last updated: 2026-10-12
- Shared benchmarks: 23

## Summary

- They share 23 benchmarks with published results for both. Claude 3 Haiku scores higher in 0 categories and Gemma 4 31B IT in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 4 31B IT leads 43.2 to 9.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.8% for Claude 3 Haiku and 73.3% for Gemma 4 31B IT.
- Gemma 4 31B IT has downloadable open weights; the other is API-only.

## Snapshot

| | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| Provider | Anthropic | Google |
| Noometry Index | 25.9 | 43.5 |
| Rank | 340 | 90 |
| Context | — | 262K |
| Input $/M | — | $0.09 |
| Output $/M | — | $0.34 |
| Weights | Proprietary | Open |

## Coding

- Claude 3 Haiku: 26.4 (#325)
- Gemma 4 31B IT: 42.3 (#108)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| WeirdML | 9.8% | 52.3% |
| LMArena Coding | 1199 | 1459 |
| LMArena WebDev | — | 1366 |
| SciCode | — | 43.4% |
| BigCodeBench Instruct | 39.4% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| ALE-Bench | — | 925.5 |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |

## Reasoning

- Claude 3 Haiku: 16.3 (#307)
- Gemma 4 31B IT: 27.2 (#122)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| Kagi LLM Benchmark | 34.2% | 63.5% |
| LMArena Hard Prompts | 1174 | 1448 |
| DTBench | 50.1% | 82.7% |
| LMCA | 8.8% | 39.3% |
| Epoch Capabilities Index | 118.35 | 142.74 |
| NYT Connections (extended) | — | 70.6% |
| CritPt | — | 1.4% |
| Chess Puzzles | — | 5% |
| Thematic Generalization | — | 53% |
| Surface Evolver Bench | — | 30.6% |
| ForecastBench | 53.2 | — |
| WinoGrande | 74.2% | — |

## Math

- Claude 3 Haiku: 9.8 (#319)
- Gemma 4 31B IT: 43.2 (#81)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.8% | 73.3% |
| LMArena Math | 1188 | 1465 |
| MATH Level 5 | 14.9% | — |

## Knowledge

- Claude 3 Haiku: 17.3 (#285)
- Gemma 4 31B IT: 37.9 (#151)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| GPQA Diamond | 36.3% | 75.8% |
| LMArena Expert | 1148 | 1465 |
| SimpleQA Verified | — | 10.4% |
| Confabulations | 34.2% | — |
| Vectara Hallucination Rate | — | 7.4% |
| MMLU | 73.8% | — |

## Multimodal

- Claude 3 Haiku: 23.6 (#128)
- Gemma 4 31B IT: 41.6 (#34)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| LMArena Vision | 950 | 1277 |
| LMArena Document | — | 1425 |
| ScienceQA | 72% | — |

## Multilingual

- Claude 3 Haiku: 36.0 (#243)
- Gemma 4 31B IT: 53.8 (#57)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| LMArena Non-English | 1178 | 1431 |
| LMArena Chinese | 1155 | 1476 |
| LMArena French | 1195 | 1435 |
| LMArena Russian | 1204 | 1460 |
| LMArena Spanish | 1166 | 1444 |
| LMArena German | 1174 | — |
| LMArena Japanese | 1102 | — |
| LMArena Korean | 1109 | — |

## Instruction Following

- Claude 3 Haiku: 61.3 (#247)
- Gemma 4 31B IT: 75.5 (#61)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| LMArena Instruction Following | 1173 | 1433 |

## Long Context

- Claude 3 Haiku: 36.1 (#237)
- Gemma 4 31B IT: 44.2 (#71)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| LMArena Longer Query | 1190 | 1446 |

## Writing & Preference

- Claude 3 Haiku: 29.7 (#291)
- Gemma 4 31B IT: 60.5 (#96)

| Benchmark | Claude 3 Haiku | Gemma 4 31B IT |
|---|---|---|
| LMArena Text | 1195 | 1443 |
| LMArena Creative Writing | 1157 | 1415 |
| EQ-Bench Creative Writing | 717 | 1368 |
| LMArena Multi-Turn | 1190 | 1452 |
| EQ-Bench 4 | — | 1120 |

## FAQ

### Is Claude 3 Haiku better than Gemma 4 31B IT?

Gemma 4 31B IT is the stronger model overall, scoring 43.5 to 25.9 on the Noometry Index.

### Is Claude 3 Haiku or Gemma 4 31B IT better for coding?

Gemma 4 31B IT scores higher on coding benchmarks: 42.3 versus 26.4 in the Noometry coding category.

### How many benchmarks do Claude 3 Haiku and Gemma 4 31B IT share?

23 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Gemma 4 31B IT has 35.