Jais-3-MoVA-36B-A4B: benchmark results
Provider: Inception (G42), Cerebras & MBZUAI. Released 2026-10-15. Access: Open.
Aggregate Rating 1704 ± 39, rank #327 of 10916 rated models, from 39 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| Jais 3 Model Cards - MCPMark | 69.3 | Pass@1 (%) | 100 |
| Jais 3 Model Cards - QIAS-1 (IFM-audited) | 95.7 | Accuracy (%) | 96.4 |
| Jais 3 Model Cards - Arabic IFEval | 80 | Accuracy (%) | 94.6 |
| Jais 3 Model Cards - BrowseComp-Plus | 71.8 | Accuracy (%) | 94.4 |
| Jais 3 Model Cards - GAIA v2 lit (Arabic) | 51.5 | Pass@1 (%) | 93.5 |
| Jais 3 Model Cards - Arabic IFBench (IFM) | 92 | Prompt-level strict accuracy (%) | 92.9 |
| Jais 3 Model Cards - IslamicEval 2025-1B (IFM-audited) | 82.7 | Accuracy (%) | 92.9 |
| Jais 3 Model Cards - QIAS-2 (IFM-audited) | 87.2 | Accuracy (%) | 92.9 |
| Jais 3 Model Cards - SimpleQA Verified (Arabic) - Non-hallucination Rate | 83.1 | Non-hallucination rate (%) | 92.9 |
| Jais 3 Model Cards - Alyah (IFM-audited) | 83 | Accuracy (%) | 91.1 |
| Jais 3 Model Cards - Arabic Cuisine (IFM) | 88.5 | Accuracy (%) | 89.3 |
| Jais 3 Model Cards - FannOrFlop MCQ | 71.5 | Accuracy (%) | 89.3 |
Interactive version: theaggregate.ai/model?slug=jais-3-mova-36b-a4b · How It Works · Data refreshed daily, snapshot 2026-11-08.