Saka-14B: benchmark results
Sakalti's community TIES merge of Qwen 2.5 14B fine-tunes. Provider: Sakalti. Released 2025-02-05. Access: Open.
Aggregate Rating 1634 ± 25, rank #344 of 1656 rated models, from 179 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Arabic LLM - Arabic MMLU Arabic Language (Grammar) | 71.23 | Accuracy (%) | 98.8 |
| Open LLM Leaderboard - GPQA | 39.6 | Score | 98.8 |
| Open Arabic LLM - Arabic MMLU HT Security Studies | 75.1 | Accuracy (%) | 98.5 |
| Open Arabic LLM - Arabic MMLU Geography (Middle School) | 75 | Accuracy (%) | 98.1 |
| Open Arabic LLM - Arabic MMLU HT Global Facts | 53 | Accuracy (%) | 97.5 |
| Open Arabic LLM - Arabic MMLU Social Science (Middle School) | 60.17 | Accuracy (%) | 97.2 |
| Open Arabic LLM - Aratrust MentalHealth | 96.05 | Accuracy (%) | 97.2 |
| Open LLM Leaderboard - MuSR | 48.86 | Score | 96.7 |
| Open Japanese LLM - CG | 64.46 | Score (%) | 96.6 |
| Open Arabic LLM - Arabic MMLU Computer Science (Middle School) | 96.3 | Accuracy (%) | 96.3 |
| Open LLM Leaderboard - MMLU-Pro | 53.96 | Score | 96 |
| Open Arabic LLM - Arabic MMLU Arabic Language (High School) | 46.92 | Accuracy (%) | 95.1 |
Interactive version: theaggregate.ai/model?slug=saka-14b · How It Works · Data refreshed daily, snapshot 2026-10-20.