Karnak: benchmark results
Arabic-English model from Egypt's Applied Innovation Center (AIC), part of the Ministry of Communications and Information Technology. Extends Qwen3-30B-A3B-Instruct-2507 to about 40B parameters, with additional Arabic-focused training and an optimized tokenizer; weights are released under Apache 2.0. Provider: Applied Innovation Center. Released 2026-02-06. Access: Open.
Aggregate Rating 1688 ± 49, rank #407 of 10916 rated models, from 133 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Open Arabic LLM - Arabic MMLU Arabic Language (General) | 81.05 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Arabic Language (Grammar) | 83.56 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Arabic Language (High School) | 60.26 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Average | 79.13 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Biology (High School) | 70.48 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Civics (Middle School) | 71.61 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Driving Test | 81.5 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Economics (High School) | 77.22 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU Economics (University) | 75.18 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU General Knowledge | 88.66 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU General Knowledge (Middle School) | 84.88 | Accuracy (%) | 100 |
| Open Arabic LLM - Arabic MMLU General Knowledge (Primary School) | 83.33 | Accuracy (%) | 100 |
Interactive version: theaggregate.ai/model?slug=karnak · How It Works · Data refreshed daily, snapshot 2026-11-08.