Gemma 4 26B A4B (Reasoning) — benchmark results
Provider: Google. Released 2026-04-02. Access: Open.
Unified ELO 1573 ± 21, rank #612 of 1839 rated models, from 35 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA IFBench | 72.45 | Accuracy (%) | 89.3 |
| AA Humanity's Last Exam | 18.26 | Accuracy (%) | 76.5 |
| Epoch AI - Scicode | 40.05 | Score | 72.7 |
| AA GPQA Diamond | 79.19 | Accuracy (%) | 70.8 |
| Artificial Analysis Intelligence Index | 25.69 | Intelligence Index | 68.6 |
| AA Long Context Reasoning | 55.67 | Accuracy (%) | 64.6 |
| AA Omniscience - Science, Engineering & Mathematics | 30.1 | Accuracy (%) | 64 |
| AA Omniscience - Business | 18.6 | Accuracy (%) | 58.6 |
| AA MMMU-Pro | 69.25 | Accuracy (%) | 55.7 |
| AA Omniscience - Software Engineering (SWE) - Swift | 40 | Accuracy (%) | 55.2 |
| Tau3 Banking | 11.75 | Success Rate (%) | 53.6 |
| AA Terminal-Bench Hard | 13.64 | Accuracy (%) | 50.1 |
Interactive version: theaggregate.ai/model?slug=gemma-4-26b-a4b-reasoning · How It Works · Data refreshed daily, snapshot 2026-08-05.