Gemma 4 12B: benchmark results
Provider: Google. Released 2026-04-02. Access: Open.
Unified ELO 1582 ± 1, rank #314 of 1392 rated models, from 24 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AI for Education Pedagogy - Technology | 85.85 | Accuracy (%) | 89.5 |
| SEA-HELM | 65.23 | Mean Score (%) | 81.4 |
| BenchmarkList ECI | 120.45 | Capability Index (ECI) | 63.3 |
| AI for Education SEND | 77.06 | Accuracy (%) | 59.5 |
| AI for Education Pedagogy - Primary | 85.45 | Accuracy (%) | 59.2 |
| AI for Education Pedagogy - Secondary | 81.29 | Accuracy (%) | 58.8 |
| ZeroEval GPQA Diamond | 78.8 | GPQA Diamond Score | 58.3 |
| AI for Education Pedagogy | 81.65 | Accuracy (%) | 57.8 |
| AI for Education Pedagogy - Maths | 80.16 | Accuracy (%) | 56.9 |
| AI for Education Pedagogy - Science | 82.51 | Accuracy (%) | 56.3 |
| LLM Stats (MathVision) | 79.7 | Score (%) | 55.9 |
| AI for Education Pedagogy - Social studies | 78.18 | Accuracy (%) | 53.6 |
Interactive version: theaggregate.ai/model?slug=gemma-4-12b · How It Works · Data refreshed daily, snapshot 2026-09-05.