Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Knowledge Evaluation on ArabicMMLU
Loading...
81.23
Accuracy
Karnak
66.7948
70.5424
74.29
78.0376
Mar 17, 2026
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
Karnak
Size=40B, Protocol=3-shot
2026.03
81.23
Jais-2-70B-Chat
Size=70B, Protocol=3-shot
2026.03
79.02
Fanar-27B
Size=27B, Protocol=3-shot
2026.03
74.67
AceGPT-v2-70B-Chat
Size=70B, Protocol=3-shot
2026.03
73.87
Llama-3.3-70b-Instruct
Size=70B, Protocol=3-shot
2026.03
73.66
Qwen3-32B
Size=32B, Protocol=3-shot
2026.03
73.08
Gemma-3-27B-it
Size=27B, Protocol=3-shot
2026.03
72.21
Allam-7B-Instruct-preview-v2
Size=7B, Protocol=3-shot
2026.03
70.01
AceGPT-v2-32B-Chat
Size=32B, Protocol=3-shot
2026.03
69.55
Fanar-1-9b-instruct
Size=9B, Protocol=3-shot
2026.03
67.35
Feedback
Search any
task
Search any
task