AI models tested for Arabic
Does it actually work for Saudi, Najdi, Hijazi or Gulf Arabic? Benchmarks rarely say. The community tests these models on real speech and text — and reports back.
AIN
MBZUAI
Bilingual Arabic-English multimodal model built on Qwen2-VL and trained on 3.6M multimodal samples, roughly a third of them authentic Arabic. Strong on Arabic OCR and document understanding, with human evaluation across 17 Arab countries. MIT-licensed.
NAMAA Saudi TTS
NAMAA Community
Saudi-dialect Arabic text-to-speech with reference-audio voice prompting, aimed at everyday conversational speech rather than Modern Standard Arabic. Notable as the only open-weights model in this directory whose card explicitly claims Saudi dialect — no model anywhere was found claiming Najdi or Hijazi specifically. MIT-licensed.