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.
TII's hybrid Transformer–Mamba series and its current open-weight flagship, released in sizes from 0.5B to 34B. Unlike Falcon 3, the model card lists Arabic among its 18 supported languages, making this the largest openly downloadable TII model with declared Arabic support.
Audar ASR V1 Turbo
Audar AI Labs
2.35B Arabic-first speech recognition trained on 300k+ hours, naming Gulf, Egyptian, Levantine and Maghrebi coverage plus Arabic-English code-switching. Its card claims the top place on the Open Universal Arabic ASR Leaderboard — a leaderboard run by Elm, a Saudi company. Custom AudarAI Community Licence, not open source.
Mistral 7B
Mistral AI
Efficient open-weight model common in regional stacks; v0.3 is the current 7B instruct release and remains Apache 2.0. Arabic is not a declared language on the card. Note that Mistral's larger flagships moved to the non-open Mistral Research Licence, so the permissive terms apply to this size, not the family.
DeepSeek-R1
DeepSeek
Open reasoning model under a plain MIT licence, allowing commercial use and derivatives including distillation. Arabic is not a declared target language — the card lists no languages — so Arabic behaviour is incidental. Superseded within the DeepSeek line by the R1-0528 update and later V3.x releases.
Qwen3
Alibaba
Open LLM family. The only model in this directory whose developer explicitly names a Saudi dialect: Qwen documents 119 languages and dialects including Arabic (Standard, Najdi, Levantine, Egyptian and others). Apache-2.0.
Llama 4
Meta
Meta's open-weight frontier family and a widely used base for regional Arabic fine-tunes. Arabic is one of the 12 languages Meta lists for Llama 4. The licence is a custom community licence with acceptable-use and above-700M-MAU conditions, not an OSI licence.
Whisper large-v3
OpenAI
State-of-the-art open speech recognition covering 99+ languages and the default open ASR baseline for Arabic products, though Arabic is one language among many rather than a focus and dialectal Arabic remains its weak point. Note the licence inconsistency: this model card states Apache 2.0 while the openai/whisper GitHub repo and the large-v3-turbo card state MIT.
Aya Expanse 32B
Cohere Labs
Open-weight multilingual model covering 23 languages including Arabic. The licence is non-commercial (CC-BY-NC-4.0) plus Cohere's acceptable use addendum, so it cannot be used in commercial Arabic products.