Gemma 3n E4B Instruct: benchmark results
Google's Gemma 3n E4B Instruct, an instruction-tuned, image-text-to-text model fine-tuned from Gemma 3n E4B. Provider: Google. Released 2025-06-26. Access: Open.
Aggregate Rating 1389 ± 19, rank #1514 of 2168 rated models, from 61 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA Omniscience - Software Engineering (SWE) - Julia | 8 | Accuracy (%) | 46.1 |
| AA Omniscience - Software Engineering (SWE) - Swift | 28 | Accuracy (%) | 40.1 |
| LLM Stats (HiddenMath) | 37.7 | Score (%) | 37.5 |
| AA Omniscience - Software Engineering (SWE) - Java | 12 | Accuracy (%) | 33.1 |
| AA Omniscience - Software Engineering (SWE) - Kotlin | 12 | Accuracy (%) | 31.7 |
| AA MATH-500 | 77.07 | Accuracy (%) | 30 |
| LLM Stats (WMT24++) | 50.1 | Score (%) | 24.1 |
| AA CritPt | 0 | Accuracy (%) | 22 |
| AA Humanity's Last Exam | 4.49 | Accuracy (%) | 20.3 |
| AA Omniscience - Software Engineering (SWE) - TypeScript | 8.89 | Accuracy (%) | 19.8 |
| AA Terminal-Bench Hard | 2.27 | Accuracy (%) | 19.6 |
| AA-Omniscience Index - Software Engineering (SWE) - Swift | -44 | Omniscience Index | 18.7 |
Interactive version: theaggregate.ai/model?slug=gemma-3n-e4b-instruct · How It Works · Data refreshed daily, snapshot 2026-10-20.