Arabic Language Evaluation on Nahw-MCQ
53.12AccuracyJais-2-70B-Chat
Evaluation Results
| Method | Links | |
|---|---|---|
| Jais-2-70B-ChatSize=70B, Evaluation Protocol=3-shot2026.03 | 53.12 | |
| Allam-7B-Instruct-preview-v2Size=7B, Evaluation Protocol=3-shot2026.03 | 51.34 | |
| Fanar-27BSize=27B, Evaluation Protocol=3-shot2026.03 | 46.88 | |
| KarnakSize=40B, Evaluation Protocol=3-shot2026.03 | 46.08 | |
| Gemma-3-27B-itSize=27B, Evaluation Protocol=3-shot2026.03 | 43.82 | |
| AceGPT-v2-70B-ChatSize=70B, Evaluation Protocol=3-shot2026.03 | 42.98 | |
| Qwen3-32BSize=32B, Evaluation Protocol=3-shot2026.03 | 42.14 | |
| Llama-3.3-70b-InstructSize=70B, Evaluation Protocol=3-shot2026.03 | 41 | |
| Fanar-1-9b-instructSize=9B, Evaluation Protocol=3-shot2026.03 | 40 | |
| AceGPT-v2-32B-ChatSize=32B, Evaluation Protocol=3-shot2026.03 | 38.62 |