Writing

Every word you write,
checked.

WriteX is Lisan's AI writing assistant: the deepest linguistic proofreader for Arabic, with English and Spanish alongside. It catches grammar, spelling, style, and clarity issues as you type, wherever you write.

Real product UI WriteX: an English sentence with a spelling and a grammar flag, the fix offered inline, category tallies and an accuracy score beside it

Arabic checked properly

Not spellcheck with an Arabic dictionary: real grammatical analysis of Arabic morphology and syntax, catching the errors that embarrass official documents.

Twelve kinds of errors

Grammar, spelling, punctuation, style, clarity, tone, and more, each explained so writers learn instead of just accepting corrections.

Three languages, one assistant

Arabic first, English in depth, and Spanish through purpose-built correction models that joined WriteX with the Correcto acquisition.

Institutional grade

More than 130 institutions run WriteX, from ministries to newsrooms. Saudi Energy runs a tailored copy for official reports and correspondence.

Everywhere you write

The web editor, browser integration, and Office add-ins bring the same engine to documents, emails, and content systems.

Faster official writing

Writers save up to 45% of their writing time in our internal measurements, because review cycles stop bouncing on language issues.

Measured

L3, the engine behind WriteX, against
26 large language model configurations.

In September 2026 Lisan Research published the evaluation of L3, the Lisan language engine that checks Arabic inside WriteX. On a linguist-reviewed benchmark of real Arabic sentences, scored on exact spans with one output per system, L3 ranked first on every primary metric: 80.29% edit F0.5 against 44.68% for Claude Fable 5.1, 84.54% precision with 66.85% recall, and 96.38% correct replacements after Arabic normalization on the spans it flagged correctly. It was second on throughput, behind Qwen 3.8 27B on Groq, and its measured energy on an NVIDIA A10 was approximately 0.000014 kWh per 1,000 tokens at accelerator scope. The report names every competitor, publishes the prompt and the scoring, and states its limits.

L3 leads on edit F0.5 at 80.29%, against 44.68% for Claude Fable 5.1

Edit F0.5, exact-span detection, top ten configurations. Higher is better. F0.5 weights precision above recall, because unnecessary changes cost more than missed ones.

Exact-span detection, micro-averaged, single output per system. Every competitor score comes from our own run of that provider's service. Measured, evaluation snapshot 9 September 2026.

Who it is for

Where WriteX fits.

Government communicationsOfficial letters, reports, and public content that must be linguistically flawless.
Newsrooms and publishersHigh-volume Arabic content with consistent style at deadline speed.
Anyone writing Arabic professionallyFrom students to executives: writing that reads the way you intended.

The full story lives on the WriteX site.

Features in depth, honest comparisons, pricing, and guides: WriteX has a dedicated site for all of it.

Start with one product.
Keep the whole platform.

Open a free account and solve today's problem in the next ten minutes. When you are ready for more, six flagships and a 20+ app workspace are already on your account. Or talk to us and we will scope it with you.

Free to start, no card · Your data exportable, always · Trusted by more than 24 government entities