jais-family-30B-16k-chat: benchmark results
Inception (G42) and Cerebras' Jais Family 30B 16K Chat, fine-tuned from their own Jais Family 30B 16K base model for a 16K-token context, released under Apache 2.0. Provider: Inception (G42) & Cerebras. Released 2024-08-02. Access: Open.
Aggregate Rating 1507 ± 21, rank #758 of 1635 rated models, from 260 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| HELM Arabic - Aratrust - Offensive | 95.65 | Exact Match (%) | 93.8 |
| Open Arabic LLM - Aratrust Offensive | 94.2 | Accuracy (%) | 93.5 |
| Open Arabic LLM - Madinah QA Average | 66.26 | Accuracy (%) | 85.8 |
| Open Arabic LLM - Arabic MMLU HT Public Relations | 63.64 | Accuracy (%) | 85.2 |
| HELM Arabic - Alghafa - Multiple Choice Facts Truefalse Balanced Task | 94.67 | Exact Match (%) | 85 |
| Open Arabic LLM - Madinah QA Arabic Language (Grammar) | 56.71 | Accuracy (%) | 83.3 |
| Open Arabic LLM - Arabic MMLU Islamic Studies | 56.96 | Accuracy (%) | 82.1 |
| Open Arabic LLM - Arabic MMLU HT Human Aging | 65.92 | Accuracy (%) | 81.5 |
| Open Arabic LLM - Arabic MMLU HT World Religions | 71.35 | Accuracy (%) | 81.5 |
| Open Arabic LLM - Alghafa Multiple Choice Facts Truefalse Balanced Task | 92 | Accuracy (%) | 81.2 |
| Open Arabic LLM - Arabic MMLU History (Middle School) | 67 | Accuracy (%) | 79.6 |
| Open Arabic LLM - Arabic MMLU Arabic Language (Grammar) | 46.58 | Accuracy (%) | 77.8 |
Interactive version: theaggregate.ai/model?slug=jais-family-30b-16k-chat · How It Works · Data refreshed daily, snapshot 2026-11-05.