# Gemma 4 26B A4B IT vs GPT-4

> Gemma 4 26B A4B IT is the stronger model overall, scoring 43.1 to 28.9 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gemma-4-26b-a4b-it-vs-gpt-4
- Last updated: 2026-10-10
- Shared benchmarks: 22

## Summary

- They share 22 benchmarks with published results for both. Gemma 4 26B A4B IT scores higher in 8 categories and GPT-4 in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemma 4 26B A4B IT leads 45.7 to 10.1.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 82.2% for Gemma 4 26B A4B IT and 1.1% for GPT-4.
- Gemma 4 26B A4B IT is cheaper at $0.0765 / $0.26 per million input/output tokens, against $30 / $60 for GPT-4.
- Gemma 4 26B A4B IT accepts more context: 262K tokens versus 8K.
- Gemma 4 26B A4B IT has downloadable open weights; the other is API-only.

## Snapshot

| | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| Provider | Google | OpenAI |
| Noometry Index | 43.1 | 28.9 |
| Rank | 88 | 308 |
| Context | 262K | 8K |
| Input $/M | $0.0765 | $30 |
| Output $/M | $0.26 | $60 |
| Weights | Open | Proprietary |

## Coding

- Gemma 4 26B A4B IT: 38.8 (#156)
- GPT-4: 31.2 (#277)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| WeirdML | 35.2% | 12.4% |
| LMArena Coding | 1446 | 1254 |
| LMArena WebDev | 1359 | — |
| SciCode | 40% | — |
| BigCodeBench Instruct | — | 46% |
| BigCodeBench Complete | — | 57.2% |
| ALE-Bench | 927.17 | — |
| HumanEval+ | — | 79.3% |

## Agentic & Tool Use

- Gemma 4 26B A4B IT: —
- GPT-4: —

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| METR Time Horizons | — | 36.1% |

## Reasoning

- Gemma 4 26B A4B IT: 21.1 (#205)
- GPT-4: 17.3 (#285)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| Chess Puzzles | 6% | 4% |
| LMArena Hard Prompts | 1438 | 1241 |
| DTBench | 74.9% | 62.7% |
| LMCA | 29.7% | 17.1% |
| Epoch Capabilities Index | 141.85 | 125.88 |
| CritPt | 0% | — |
| Mystery Game Puzzles | — | 12% |
| BIG-Bench Hard | — | 75.1% |
| ForecastBench | — | 57.8 |
| HellaSwag | — | 95.3% |
| WinoGrande | — | 87.5% |

## Math

- Gemma 4 26B A4B IT: 45.7 (#61)
- GPT-4: 10.1 (#304)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 82.2% | 1.1% |
| LMArena Math | 1469 | 1269 |
| MATH Level 5 | — | 23% |
| GSM8K | — | 92% |

## Knowledge

- Gemma 4 26B A4B IT: 45.6 (#79)
- GPT-4: 18.4 (#277)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| GPQA Diamond | 73.2% | 35.7% |
| LMArena Expert | 1447 | 1211 |
| Vectara Hallucination Rate | 5.2% | — |
| MMLU | — | 86.4% |
| TriviaQA | — | 84.8% |

## Multimodal

- Gemma 4 26B A4B IT: 39.9 (#43)
- GPT-4: —

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| LMArena Vision | 1260 | — |

## Multilingual

- Gemma 4 26B A4B IT: 53.3 (#68)
- GPT-4: 40.8 (#210)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| LMArena Non-English | 1422 | 1246 |
| LMArena Chinese | 1495 | 1242 |
| LMArena French | 1459 | 1282 |
| LMArena Russian | 1434 | 1251 |
| LMArena Spanish | 1415 | 1261 |
| LMArena German | — | 1251 |
| LMArena Japanese | — | 1209 |
| LMArena Korean | — | 1184 |

## Instruction Following

- Gemma 4 26B A4B IT: 75.0 (#78)
- GPT-4: 65.5 (#217)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| LMArena Instruction Following | 1420 | 1241 |

## Long Context

- Gemma 4 26B A4B IT: 43.6 (#87)
- GPT-4: 37.7 (#207)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| LMArena Longer Query | 1427 | 1244 |

## Writing & Preference

- Gemma 4 26B A4B IT: 58.8 (#110)
- GPT-4: 35.0 (#263)

| Benchmark | Gemma 4 26B A4B IT | GPT-4 |
|---|---|---|
| LMArena Text | 1434 | 1263 |
| LMArena Creative Writing | 1401 | 1244 |
| EQ-Bench Creative Writing | 1305 | 752 |
| LMArena Multi-Turn | 1442 | 1257 |

## FAQ

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

Gemma 4 26B A4B IT is the stronger model overall, scoring 43.1 to 28.9 on the Noometry Index.

### Which is cheaper, Gemma 4 26B A4B IT or GPT-4?

Gemma 4 26B A4B IT is cheaper. It lists at $0.0765 per million input tokens and $0.26 per million output tokens; GPT-4 lists at $30 and $60.

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

Gemma 4 26B A4B IT scores higher on coding benchmarks: 38.8 versus 31.2 in the Noometry coding category.

### Which has the bigger context window?

Gemma 4 26B A4B IT does, with 262K tokens against 8K.

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

22 benchmarks have published results for both models. Gemma 4 26B A4B IT has 28 scored results on Noometry and GPT-4 has 38.