Gemma 4 26B A4B (Reasoning): benchmark results
Provider: Google. Released 2026-04-02. Access: Open.
Aggregate Rating 1554 ± 21, rank #867 of 2170 rated models, from 48 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA IFBench | 72.45 | Accuracy (%) | 89.3 |
| AA Omniscience - Software Engineering (SWE) - Swift | 48 | Accuracy (%) | 82.4 |
| AA Omniscience - Software Engineering (SWE) - Dart | 24 | Accuracy (%) | 72.2 |
| AA GPQA Diamond | 79.19 | Accuracy (%) | 65.9 |
| AA Humanity's Last Exam | 19.32 | Accuracy (%) | 64.8 |
| AA Omniscience - Software Engineering (SWE) - R | 12 | Accuracy (%) | 62.1 |
| AA Omniscience - Software Engineering (SWE) - PHP | 24 | Accuracy (%) | 60.6 |
| AA Omniscience - Software Engineering (SWE) - C | 35 | Accuracy (%) | 58.4 |
| AgentVidBench (Single-Turn) | 22 | Accuracy (%; 100 human-authored multi-hop multiple-choice qu | 57.9 |
| AA Omniscience - Software Engineering (SWE) - Java | 16 | Accuracy (%) | 57 |
| AA Omniscience - Software Engineering (SWE) - JavaScript | 29.09 | Accuracy (%) | 57 |
| AA Omniscience - Software Engineering (SWE) - Go | 18 | Accuracy (%) | 56.5 |
Interactive version: theaggregate.ai/model?slug=gemma-4-26b-a4b-reasoning · How It Works · Data refreshed daily, snapshot 2026-11-05.