# Gemma 4 26B A4B IT vs GPT-4.5

> Gemma 4 26B A4B IT is the stronger model overall, scoring 43.5 to 37.2 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemma-4-26b-a4b-it-vs-gpt-4-5
- Last updated: 2026-10-12
- Shared benchmarks: 19

## Summary

- They share 19 benchmarks with published results for both. Gemma 4 26B A4B IT scores higher in 8 categories and GPT-4.5 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 4 26B A4B IT leads 47.6 to 32.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 82.2% for Gemma 4 26B A4B IT and 37.8% for GPT-4.5.
- Gemma 4 26B A4B IT has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 43.5 | 37.2 |
| Rank | 92 | 208 |
| Context | 262K | — |
| Input $/M | $0.09 | — |
| Output $/M | $0.30 | — |
| Weights | Open | Proprietary |

## Coding

- Gemma 4 26B A4B IT: 39.0 (#164)
- GPT-4.5: 42.2 (#109)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| WeirdML | 35.2% | 39.4% |
| LMArena Coding | 1447 | 1396 |
| Aider Polyglot | — | 44.9% |
| LMArena WebDev | 1359 | — |
| SciCode | 40% | — |
| LiveBench Coding | — | 75.2% |
| ALE-Bench | 927.17 | — |

## Agentic & Tool Use

- Gemma 4 26B A4B IT: —
- GPT-4.5: 27.9 (#97)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| Cybench | — | 17.5% |

## Reasoning

- Gemma 4 26B A4B IT: 21.8 (#213)
- GPT-4.5: 13.9 (#330)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| LMArena Hard Prompts | 1439 | 1403 |
| Epoch Capabilities Index | 141.85 | 136.74 |
| ARC-AGI-2 | — | 0.8% |
| SimpleBench | — | 34.5% |
| ARC-AGI-1 | — | 10.3% |
| CritPt | 0% | — |
| Chess Puzzles | 6% | — |
| EnigmaEval | — | 3.2% |
| LiveBench Reasoning | — | 71.1% |
| DTBench | 74.9% | — |
| LiveBench Data Analysis | — | 64.3% |
| LMCA | 29.7% | — |
| ForecastBench | — | 61.7 |
| LiveBench | — | 69% |

## Math

- Gemma 4 26B A4B IT: 47.6 (#67)
- GPT-4.5: 32.6 (#211)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 82.2% | 37.8% |
| LMArena Math | 1470 | 1412 |
| LiveBench Math | — | 69.3% |
| MATH Level 5 | — | 78.6% |

## Knowledge

- Gemma 4 26B A4B IT: 45.7 (#85)
- GPT-4.5: 32.5 (#211)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| GPQA Diamond | 73.2% | 68.7% |
| LMArena Expert | 1447 | 1394 |
| Humanity's Last Exam | — | 5.4% |
| Confabulations | — | 13.6% |
| Vectara Hallucination Rate | 5.2% | — |

## Multimodal

- Gemma 4 26B A4B IT: 40.6 (#46)
- GPT-4.5: 37.6 (#71)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| LMArena Vision | 1260 | 1195 |
| VPCT | — | 45% |

## Multilingual

- Gemma 4 26B A4B IT: 53.1 (#71)
- GPT-4.5: 52.5 (#83)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| LMArena Non-English | 1421 | 1413 |
| LMArena Chinese | 1495 | 1421 |
| LMArena French | 1460 | 1418 |
| LMArena Russian | 1434 | 1419 |
| LMArena German | — | 1457 |
| LMArena Japanese | — | 1416 |
| LMArena Korean | — | 1392 |
| LMArena Spanish | 1417 | — |

## Instruction Following

- Gemma 4 26B A4B IT: 74.8 (#82)
- GPT-4.5: 72.6 (#134)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| LMArena Instruction Following | 1420 | 1404 |
| LiveBench Instruction Following | — | 72.3% |

## Long Context

- Gemma 4 26B A4B IT: 43.6 (#91)
- GPT-4.5: 40.4 (#155)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| LMArena Longer Query | 1428 | 1406 |
| Fiction.LiveBench | — | 63.9% |

## Writing & Preference

- Gemma 4 26B A4B IT: 58.6 (#115)
- GPT-4.5: 56.9 (#134)

| Benchmark | Gemma 4 26B A4B IT | GPT-4.5 |
|---|---|---|
| LMArena Text | 1434 | 1417 |
| LMArena Creative Writing | 1402 | 1394 |
| EQ-Bench Creative Writing | 1305 | 1258 |
| LMArena Multi-Turn | 1441 | 1444 |
| Short-Story Creative Writing | — | 75.6% |
| LiveBench Language | — | 61.5% |

## FAQ

### Is Gemma 4 26B A4B IT better than GPT-4.5?

Gemma 4 26B A4B IT is the stronger model overall, scoring 43.5 to 37.2 on the Noometry Index.

### Is Gemma 4 26B A4B IT or GPT-4.5 better for coding?

GPT-4.5 scores higher on coding benchmarks: 42.2 versus 39.0 in the Noometry coding category.

### How many benchmarks do Gemma 4 26B A4B IT and GPT-4.5 share?

19 benchmarks have published results for both models. Gemma 4 26B A4B IT has 28 scored results on Noometry and GPT-4.5 has 42.