Skip to content
Muarrab

Developer resources

Shipping Arabic without breaking your build.

Most Arabic releases do not fail at translation. They fail at shaping, direction, line height, search, subtitle timing and store pages — the parts that are nobody's assigned job until launch week. This is the checklist for those parts.

Jump to the checklist See what actually shipped

Start from the evidence

What the existing record tells you

Before scoping anything, look at what has already been done for games like yours. Three numbers from this index shape the decision.

  • 292

    of 307 ship without Arabic voice

    Text-only localisation is the normal outcome in this market. If voice is out of scope, you are in good company — and you are not shipping nothing.

  • 60

    teams are already doing this work

    You are not creating a market. You are choosing a supplier in one that exists, and the list is public.

  • 86

    entries name no team at all

    Attribution in this field is unreliable. Being able to name who did your localisation is itself a selling point to players.

01 — Text and type

Text and type

Arabic is a different letterform system, not a character set. Most first-pass failures happen here, before a single string is translated.

  • Glyph coverage. Arabic uses contextual letterforms — a letter changes shape by position — plus ligatures and optional diacritics. A Latin-only font will drop to fallback fonts inconsistently or show tofu boxes. Ship a font that covers Arabic and audit every weight and size you actually render.

  • Shaping. Text has to be shaped at render time. Any pipeline that measures, wraps, or truncates strings by counting characters assumes one glyph per character, which is not true in Arabic.

  • Vertical room. Arabic generally needs more line height than Latin at the same size. Text that fits comfortably in English can clip its ascenders and descenders in Arabic.

  • Numerals. Decide between Western Arabic (0–9) and Arabic-Indic (٠–٩) digits and apply the decision consistently across UI, store pages and marketing. Expectations differ by market.

  • Truncation. Ellipsising or cutting mid-word can slice a connected letterform into something that reads as broken rather than shortened.

02 — Direction and layout

Direction and layout

Right-to-left is a property of a text run, not a setting on a page. Treating it as a page-level flip produces the classic RTL bugs.

  • Bidirectional text. Arabic sentences can contain Latin words, numbers, file paths and brand names — and Latin sentences can contain Arabic. Use the Unicode bidirectional algorithm rather than reversing strings by hand.

  • Mirroring is not flipping. Layout order mirrors; many individual elements should not. Logos, clocks, media transport controls, real-world maps and brand marks usually stay as they are, while navigation, progress and back/forward affordances mirror.

  • Directional controls. Sliders, progress bars, timelines, carousels and scroll indicators each have a direction. Decide explicitly what “forward” and “next” mean for each one.

  • Charts and data. Axis order, legend placement and reading order change. A graph that reads naturally in English can be near-unreadable mirrored.

  • Mixed-direction input. Player names, chat, file names and search queries can switch direction mid-string. Test the fields where users type, not only the ones you ship.

03 — Input, search and data

Input, search and data

Arabic breaks assumptions that hold quietly in Latin-script products. These are the ones that generate support tickets.

  • Search normalisation. Users type Arabic with and without diacritics and with different forms of the same letter (alef variants, ta marbuta versus ha). Normalise queries and stored text, or search silently fails.

  • Sorting and indexing. Arabic alphabetical order is not Latin codepoint order. Any A–Z sort applied to Arabic data will look random to a native reader.

  • No letter case. Arabic has no uppercase and lowercase. Uppercase transforms are no-ops at best, and can corrupt data at worst.

  • Length validation. Validate user-facing text by rendered width, not character count. Some Arabic strings are shorter than English, others are considerably longer.

  • Font fallback in user content. If users can enter text, they can enter scripts your UI font does not cover.

04 — Audio and voice

Audio and voice

Voice is the expensive decision, and it is a project of its own rather than a translation task.

  • Casting and direction. Dubbing needs native casting, an artistic director, and a performance pass in the target language — not a translated script read by whoever is available.

  • Re-timing. Subtitles, on-screen text tied to dialogue, tutorials and audio cues all need their own timing pass in Arabic. A translated subtitle file is not a finished one.

  • Lip-sync versus line length. If you are not matching lips, cast for line length so the mix still lands. Otherwise subtitles and audio drift apart.

  • Know the baseline: only 15 of 307 documented Arabic localizations include any voice work. Text-only is the norm in this market, not a compromise nobody accepts.

05 — Store, platform and certification

Store, platform and certification

The game is one localisation surface among several, and the other ones are usually forgotten until launch week.

  • Store pages and metadata. Title, short description, long description, feature bullets, tags, screenshots and trailers are separate surfaces with their own character limits. Screenshots showing English UI undermine an Arabic store page.

  • Certification requirements. Each platform states which elements must be localised and which may remain in English, and those requirements change. Check the current published requirements for each platform rather than assuming.

  • Age ratings and descriptors. Rating bodies and content descriptors differ across the region. What is acceptable in one market may need a different descriptor in another.

  • Support and legal. EULAs, privacy policies, refund flows and customer support in Arabic are part of the shipped product, not paperwork.

06 — Quality assurance

Quality assurance

Testing is where an Arabic release is actually won. Almost none of these failures are visible in a translated string table.

  • Test in the build, in context, with the shipped font at the shipped resolution. A spreadsheet review cannot see shaping, clipping or direction.

  • Pseudo-localisation does not catch bidi or shaping problems. Run a real Arabic pass from the first milestone rather than adding it before launch.

  • Use native-speaker LQA against the live build, not only against strings..

  • Know which screens break first: settings menus, long player names, error and network strings, save-slot dates and timestamps, store pages, and anything with a progress indicator..

  • Record what you shipped. Scope, platforms and the team that did the work are what your players will want to check — and what a catalog like this one asks for.

Scope decisions

Three choices that decide the budget

  • Text only, or voice?

    These are different projects. Text localisation is an engineering and translation exercise you can scope inside a milestone; voice is casting, direction, recording, mixing and re-timing across every language you ship. 15 of the 307 documented localizations in this index include any Arabic voice work, so text-only is a legitimate destination rather than a half-measure.

  • Modern Standard Arabic, or a dialect?

    MSA is the default for interface text and is understood across the region. Dialects matter most in voice work and marketing, where audiences are more sensitive to register. Whichever you choose, be consistent inside the product, and say which one the store page offers.

  • Official release, or leave it to the community?

    If a community patch already exists, the demand is proven — and so is the competition. Players may prefer a patch they know to an official version they do not, particularly if key terminology or a character's voice differ. Check what already exists before you scope yours.

Pre-flight

Hand this to your engineering lead

Twelve things to confirm before an Arabic build is considered done.

  1. 01 A font that covers Arabic, shipped and used at every weight you render.
  2. 02 Bidi handling through the platform's text stack, not manual string reversal.
  3. 03 The interface mirrored, with icons audited individually rather than flipped in bulk.
  4. 04 Line heights and container sizes re-checked in Arabic, not inherited from English.
  5. 05 Search and sorting that behave correctly for Arabic input and Arabic alphabetical order.
  6. 06 Numerals decided, applied, and consistent with your store page.
  7. 07 Every user-facing text field tested with Arabic and mixed-direction input.
  8. 08 Subtitles re-timed for Arabic, not translated and dropped in.
  9. 09 Store page, screenshots and trailer localised alongside the game.
  10. 10 Current certification requirements checked per platform.
  11. 11 A native-speaker LQA pass in the build before launch.
  12. 12 Scope, platforms and the responsible team documented — so players can verify what you shipped.

Every team that has done this work is in the index

Muarrab does not sell localization and does not rank the teams that do. What it holds is the record: who shipped Arabic, on what, and whether it was publisher-commissioned or volunteer work. Take the checklist above and ask each of them which parts they have actually delivered.

See the teams

Sources

  1. Arabic is supported by 7.8% of Steam's best-selling games; Korean reaches 49.9%. Steam's own store search filters, across 6,387 top-selling games, dated 27 Sep 2026. Source ↗
  2. MENA-3 revenue, player counts and revenue per player. Niko Partners, 2025 MENA Gamer Behavior & Market Insights, reported by PocketGamer.biz on 15 July 2025. Source ↗
  3. Wider Middle East market size, player population, platform split and growth. Analysys Mason, 20 August 2024. Source ↗
  4. Saudi market size projection, company and job targets. General Authority of Media Regulation, 2025 Media Sector Report, cited by playinarabic.com. Source ↗
  5. Gulf and Saudi player language preference. Regional player research and market surveys across GCC countries.
  6. Arab world ranks #13 of 239 in Steam global downloads (1.69%), ahead of Australia and Spain. Valve Steam Download Traffic by Country, archived 27 Sep 2024. Source ↗
  7. 72 million gamers across Saudi Arabia, UAE, and Egypt exceed 7 European countries combined. Niko Partners (2024 MENA-3 Report) and Video Games Europe (2023 European country demographics). Source ↗
  8. PlayStation Arabic localization commercial impact on game and console sales. Jim Ryan (CEO/Global Sales, Sony Interactive Entertainment), IGN Middle East interview, 2016. Source ↗

Catalog counts, voice coverage, team, genre and platform tallies — computed from Muarrab's own index on 2026-09-29. Page visits from Google Analytics (G-FVF8XZWNYQ).