Arabic AI Lawyer Apps Compared (2026): Adel, Shwra & More

Why Arabic is genuinely hard for AI

Most 'AI lawyer' apps are English models in a trench coat. That works until the language fights back, and Arabic fights back hard.

Key facts

• Arabic legal search surfaced 9× more primary law than English in HAQQ's test — but with dangerous jurisdiction-mixing errors.
• Adel (Saudi Arabia): 663 App Store ratings at 4.6★, a claimed 500K downloads, 70,000+ Saudi legal documents, SAR 149-199/month.
• No Arabic legal AI product is simultaneously consumer, mobile, multi-country, and native-Arabic — every player hits two or three of the four corners.

It's written right-to-left, which breaks naive text pipelines. It's diglossic: the Modern Standard Arabic of statutes is not the Egyptian or Gulf dialect people actually type, and legal Arabic is a third register again — formal, archaic, full of terms of art. A model fluent in conversational Arabic can still miss a contract clause the way a fluent English speaker might fumble a 17th-century deed.

And — counterintuitively — the problem usually isn't missing content. We ran the same legal questions in both languages and Arabic surfaced 9× more primary law than English; the catch was that those sources sit un-indexed on bare government servers, and retrieval kept mixing up which country's law it found. We wrote that up in full in the Arabic legal AI gap — the short version is that the hard part is retrieval and jurisdiction accuracy, not vocabulary. This post is about the layer above that: who's actually shipping products on top of this reality, and how far they've got.

Who's actually building it

More people than the English-language press thinks. Honestly mapped, by country:

Saudi Arabia — the deepest market. Adel is the most traction-proven Arabic legal AI we found: a consumer-and-pro app with 663 App Store ratings (4.6★), a claimed 500K downloads, 70,000+ Saudi legal documents, and real published pricing (SAR 149–199/month). Shwra has the widest consumer distribution — iOS, Android and Huawei, ~1,400 App Store ratings — built as a hybrid: an AI assistant ('Mishir') that triages and then routes you to a licensed human lawyer. Qaanoon is the free, no-friction Arabic chatbot for individuals. Laika and Malakah round out a genuinely competitive Saudi field. Every one of them is Saudi-only.

UAE. realLaw is a consumer UAE app (free tier + AED 74/month) carrying a Dubai government AI certification. Qanooni ($2M pre-seed, Village Global) goes the B2B route with Word/Outlook integration. And the wildcard isn't a startup at all: the UAE Ministry of Justice, with vendor GenArabia, has deployed an Arabic legal-AI assistant covering 5,000+ pieces of legislation — as kiosks at court entrances. Free, government-run, and reshaping what citizens expect.

Egypt. LegalMind, 'made by Egyptian lawyers for Egyptian lawyers,' is the dedicated Egyptian-law tool — B2B, web-only, one country.

Pan-Arab. Laiwyer.ai is the rare multi-country play — Qatar, UAE, KSA and Egypt, transparent pricing ($49–99/mo) — but it's a web-only research tool for lawyers, not a consumer app. And Arabic.ai (partnered with Jordan's Qistas) is the most technically serious of all: a genuinely sovereign Arabic-first LLM — its own models, 22 dialects plus MSA — not a wrapper on GPT. But it sells only to enterprises and governments, behind a procurement cycle. No consumer can touch it.

The four-corner test

Line them up against four questions a person — not a law firm — actually cares about:

• Can a regular consumer use it? (not just lawyers)
• Is it on my phone? (a real mobile app, not a web login)
• Does it cover more than one country's law?
• Is it natively Arabic? (RTL, dialect-aware — not English with a translate button)

Adel and Shwra ace 1, 2 and 4 — and fail 3 (Saudi only). Laiwyer aces 3 and 4 — and fails 1 and 2 (web, lawyers). Arabic.ai has the best 4 in the business — and fails 1, 2 and 3 for any individual. Every serious player lands two or three corners. None lands all four. That's not a knock on any of them; single-country depth is a perfectly good strategy. It's just where the open space is.

Every Arabic legal AI, compared

The full MENA/Arabic field we found, with the honest details — audience, platform, pricing, whether the Arabic is native, and whether it crosses borders. (We excluded Perle AI: despite the 'Arabic legal' framing, it's a data-annotation platform, not a legal tool.)

Read down the last two columns. The two 'yes-and-yes' rows that are also consumer + mobile? There's one — and it launched last week.

Three gaps nobody's closed

• Web, not mobile. The most technically advanced Arabic legal AI (Arabic.ai, Laiwyer, LegalMind, Oqood) lives in a browser tab. The phone-native experiences are the consumer Saudi apps — which brings us to gap two.
• Lawyer, not consumer. Much of the funding and the best models are aimed at law firms. The individual with a tenancy dispute in Cairo or an employment question in Dubai is an afterthought.
• One country, not many. The whole region shares a legal heritage and a language, yet almost every product stops at one border. A Lebanese founder operating in the UAE and selling into Saudi has to open three different tools.

What we're doing about it

Disclosure, since this is our blog: we build HAQQ Legal AI, and we built it to aim at all four corners at once — consumer, mobile, multi-jurisdiction, native Arabic. Our engine, Justinian, handles Arabic right-to-left as a first-class language and is jurisdiction-aware across 80+ countries' legal systems, not one. The app is live on the App Store (Android is coming).

The honest caveat, same as our global app teardown: we're the newest name on this page. Adel has half a million downloads and we have a launch. Arabic.ai has a sovereign model with benchmark scores we respect and don't claim to beat. What we have is a bet on the one combination nobody else is making — and after mapping the field, we're more convinced the corner is real than that we're the only ones who'll ever stand in it.

Key takeaways

• MENA's Arabic legal-AI scene is real and competitive — the 'nothing here' narrative is outdated.
• The strongest consumer apps are Saudi-only; the strongest model is enterprise-only; the multi-country tools are web-only and lawyer-facing.
• No single product is consumer + mobile + multi-country + native-Arabic. Every player hits two or three corners.
• That open corner is exactly what we built HAQQ for — as the newest, least-proven name on the list.

Sources & further reading

• where MENA regulators stand on lawyers using AI
• what lawyers actually charge across MENA
• The global 110-app teardown
• Why the Arabic gap is retrieval, not content
• State of Legal AI in MENA 2026
• Arabic.ai + Qistas partnership (Wamda)
• Oqood raises $1M seed (Wamda)

← All HAQQ articles