SILMA-Kashif-2B-Instruct-v1.0 — benchmark results
SILMA AI's Gemma-based 2B Arabic/English model (January 2025) specialized for RAG question answering, with 12K context. Provider: SILMA AI. Released 2025-01-26. Access: Open.
Unified ELO 1413 ± 13, rank #1165 of 1776 rated models, from 130 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Arabic LLM - Arabic MMLU LAW (Professional) | 67.83 | Accuracy (%) | 59.9 |
| Open Arabic LLM - Aratrust Ethics | 70 | Accuracy (%) | 57.1 |
| Open Arabic LLM - Aratrust Privacy | 91.23 | Accuracy (%) | 55.9 |
| Open Arabic LLM - Arabic MMLU HT Virology | 43.37 | Accuracy (%) | 53.4 |
| Open Arabic LLM - Aratrust MentalHealth | 89.47 | Accuracy (%) | 51.9 |
| Open Arabic LLM - Arabic MMLU General Knowledge (Middle School) | 59.88 | Accuracy (%) | 46 |
| Open Arabic LLM - Alghafa Multiple Choice Facts Truefalse Balanced Task | 78.67 | Accuracy (%) | 44.4 |
| Open Arabic LLM - Arabic MMLU HT High School Physics | 31.79 | Accuracy (%) | 44.1 |
| Open Arabic LLM - Aratrust Offensive | 73.91 | Accuracy (%) | 43.8 |
| Open Arabic LLM - Arabic MMLU Philosophy (High School) | 51.28 | Accuracy (%) | 43.5 |
| Open Arabic LLM - Arabic MMLU General Knowledge (Primary School) | 57.41 | Accuracy (%) | 42.9 |
| Open Arabic LLM - Aratrust PhysicalHealth | 75.34 | Accuracy (%) | 42.9 |
Interactive version: theaggregate.ai/model?slug=silma-kashif-2b-instruct-v1-0 · How the rankings work · Data refreshed daily, snapshot 2026-07-22.