gemma-3-12B-pt: benchmark results
Google's Gemma 3 12B pretrained base model - open weights with text+image input and a 128K context (March 2025). Provider: Google. Released 2025-03-12. Access: Open.
Aggregate Rating 1529 ± 15, rank #685 of 1656 rated models, from 604 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| EuroEval Czech NLU - CSFD Sentiment | 75.96 | Matthews correlation coefficient (%) | 100 |
| EuroEval Slovak NLU - CSFD Sentiment SK | 68.04 | Matthews correlation coefficient (%) | 100 |
| EuroEval Swedish NLU - Swerec | 81.27 | Matthews correlation coefficient (%) | 100 |
| EuroEval Ukrainian NLU - Cross Domain UK Reviews | 65.19 | Matthews correlation coefficient (%) | 100 |
| LatamBoard - Spanish COPA | 85.8 | Score (%) | 100 |
| LatamBoard - Spanish MGSM | 30.8 | Score (%) | 100 |
| Ukrainian LLM - ARC Easy UK | 80.64 | Score (%) | 100 |
| Ukrainian LLM - Flores UK CRH | 3.88 | Score (%) | 100 |
| Ukrainian LLM - Flores UK CS | 25.82 | Score (%) | 100 |
| Ukrainian LLM - Flores UK DE | 28.91 | Score (%) | 100 |
| Ukrainian LLM - Flores UK HU | 19.63 | Score (%) | 100 |
| Ukrainian LLM - Flores UK PL | 19.76 | Score (%) | 100 |
Interactive version: theaggregate.ai/model?slug=gemma-3-12b-pt · How It Works · Data refreshed daily, snapshot 2026-10-21.