▸ multilingual niche
What is the best LLM
for Arabic?
A Arabic winner is published only when both independent per-language sources match callable models. This is a cross-source claim: Global-MMLU-Lite and the AI Language Proficiency Monitor must both report the model in Arabic.
▸ evidence used
Global-MMLU-Lite Arabicindependent multilingual knowledge result
Language Proficiency Monitor ArabicARC + MGSM + MMLU accuracy panel
▸ what this rank does not claim
This score tests multilingual knowledge and reasoning in Arabic. It does not prove tone, dialect quality, translation quality, or legal accuracy.
Models missing either required source are omitted rather than assigned a flattering score from incomplete evidence.