gemma-3-12B-Turkish-V1: benchmark results
Provider: Google. Access: API.
Aggregate Rating 1590 ± 41, rank #1118 of 10916 rated models, from 21 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| TurkBench - STS | 92.55 | Pearson correlation x100 | 81.1 |
| TurkBench - Named Entity Recognition | 63.88 | Accuracy (%) | 75.7 |
| TurkBench - Metaphors and Idioms | 74.67 | Accuracy (%) | 71.6 |
| TurkBench - Summarization | 76.99 | LLM-as-a-judge score (0-100) | 70.3 |
| TurkBench - Turkish Vocabulary | 90 | Accuracy (%) | 68.9 |
| TurkBench - Commonsense Reasoning | 90.98 | Accuracy (%) | 63.5 |
| TurkBench - Sentiment Analysis | 46.72 | Accuracy (%) | 60.8 |
| TurkBench - Turkish General Knowledge | 71.35 | Accuracy (%) | 58.1 |
| TurkBench - Complex Reasoning | 26.92 | Accuracy (%) | 55.4 |
| TurkBench - NLI | 71.37 | Accuracy (%) | 55.4 |
| TurkBench - Parts of Speech | 73.44 | Accuracy (%) | 54.1 |
| TurkBench - Reading Comprehension | 92.59 | LLM-as-a-judge score (0-100) | 54.1 |
Interactive version: theaggregate.ai/model?slug=gemma-3-12b-turkish-v1 · How It Works · Data refreshed daily, snapshot 2026-11-08.