110 languages, and all four services in every one

GoLocalise works in 110 languages, and offers all four services — voice over, subtitling, transcription and translation — in every one of them. The full list is the table on this page, so the claim can be checked rather than taken. Regional variants are listed but not counted twice: the seven English voice pages are one language, not seven, and where a service names the same language differently — Filipino and Tagalog, French and French (European) — it is still one row. Of those 110, 99 have a dedicated voice over page written, 110 a transcription page, 88 a subtitling page and 87 a translation page; the rest are covered but not yet written up.

GoLocalise can. We are a London voice over and localisation company, recording since 2008 from two central-London studios, with over 1,000 voice actors on the books. We record voice over in 110 languages, of which 99 have a dedicated page written so far, across 108 pages once regional variants are counted separately. The roster also covers languages that are not in the table at all. The full list is on this page, each one linked either to its page or straight to a quote.

No. The table lists the languages we have written up, which is what can be verified from the outside. We regularly work in languages that are not on it, and where we do not already have someone we go and find them: sourcing native speakers for the language and the accent, testing them on material like yours, and having a second native check the work before any of it reaches you. That takes longer than a job in a language we already staff, and for a language with a genuinely small professional pool it is not always possible — but you get a real answer either way rather than a maybe. Tell us the language and the service and we will say whether we have people for it now or need to recruit for it, and roughly how long that takes.

The language you get back. Every row is a deliverable, and what you send is a separate question that can be any language on the list. It matters most for subtitling and translation, where the two ends differ: a Cantonese video you want subtitled into English is an English subtitling job, and an English video you want subtitled for Hong Kong is a traditional-character job — same Chinese, two different rows, because you are buying two different things. All four hand you a file, and which file is worth being exact about: voice over an audio file; transcription a transcript, with timecodes or without; subtitling an SRT or another subtitle file, which is itself a text file; translation your own document translated — the Word, the Excel, the script, in the format you sent it. What changes between them is what decides the language that file comes out in. In voice over and transcription the audio decides it: you record in the language being spoken, and you transcribe what is on the tape, so naming those services by the spoken language names the file too. In subtitling and translation you decide it, because the file comes out in a target language that need not be the source at all. In practice you never have to work this out — tell us what you have and what you need, and we will tell you which pair that is.

A red tick links to a page written about that language and that service. A grey tick means nothing is written for that combination yet, so it goes to a quote form instead. Two cells are not ticks at all, and they carry the most useful thing on the page. A variant row reading 'as English' means the variant makes no difference to that service: an accent changes who records a voice over, so each English accent has its own voice over page, but it changes nothing in writing — there is no British English subtitling as a service separate from English subtitling, so one page covers every variant. A parent row reading 'by variant' is the same fact from the other side: English has no general voice over page because the work is always a specific accent, and those pages are the rows underneath. Chinese works on a second axis and the table shows it directly. Voice over and transcription name what is spoken — Mandarin, Cantonese — while subtitling and translation name what is written, because that is the deliverable. Which characters depends on the market: mainland China and Singapore use simplified, Taiwan and Hong Kong traditional. So the Cantonese row reads 'in Traditional' for subtitling, because a Hong Kong video is subtitled in traditional characters even though nobody speaks 'traditional'; Mandarin reads 'in Simplified', and the Taiwanese Mandarin row underneath reads 'in Traditional'. Translation has the fewest pages written so far, which is why that column has the most grey.

Variants, where the market difference is real. English is recorded in British, American, Australian, Irish, New Zealand, South African and African varieties; Portuguese in European and Brazilian; Chinese in Mandarin and Taiwanese, with Cantonese treated as its own language rather than a variant, because it is one. Where a variant has a page of its own it appears in the table indented under its parent language.

Yes, and you should. Each voice over page has a library you can filter and play, and for anything you are seriously considering we will record a free casting on your own script rather than send you a generic showreel — a demo tells you what an actor sounds like reading someone else's words, which is not the same question. For the smaller languages there is no public demo wall on purpose: the professional pool is genuinely small, so a fixed list would be misleading about who is free in any given week. There we cast fresh against your brief.

Yes, and the handling matters more than the recording does. Right-to-left text mixes with left-to-right runs — brand names, numbers, URLs, product codes — and the order those appear in is decided by the Unicode bidirectional algorithm, not by whoever typed the file. Text that looks correct in one editor can render scrambled in a player or a burned-in subtitle. We build and check the file in the destination format rather than converting a generic one at the end, which is where most right-to-left jobs go wrong.

They need different limits, and the limits are per language rather than per script. Japanese subtitles run to about 13 characters a line at roughly 4 characters a second; Simplified Chinese 16 characters at 9; Thai 35 characters excluding the tone marks and vowels that sit above and below the line, which do not count. For comparison, Spanish takes 42 characters at 17. Applying one language's figure to another is one of the commonest ways to get a file rejected in QC, so each one is checked against its own specification.

That is most of what we do. Send one video or script and we can run transcription, translation, subtitling, voice casting, recording and final delivery across as many languages as you need, in parallel, with one project manager, one quote and one deadline. The practical advantage is not the price — it is that terminology, timings and the master file stay consistent across all of them, which stops being true the moment different suppliers handle different languages.

It depends on the market, and it is worth getting right because audiences notice. Spanish for Latin America is a separate recording from Spanish for Spain, and Canadian French is a separate market with a legal dimension rather than an accent of French. For written Chinese, mainland China and Singapore use Simplified, Taiwan uses Traditional, and Hong Kong speaks Cantonese while writing Traditional — so a Cantonese video is normally subtitled in Traditional characters, not in written Cantonese. Tell us the market rather than the language and we will tell you which variant that actually means.

Either. Most clients send the original and we handle the translation and adaptation as part of the job, which is usually better because a script translated without knowing it will be spoken, or subtitled, reads wrong when it gets there — spoken copy needs to fit the timing and subtitled copy needs to fit the reading speed. If you already have a translation, we will work from it, and we will tell you if it needs adapting before it goes to a voice or on screen rather than quietly recording something that does not fit.

Yes. Every language in this table is worked by native speakers, and voice over is recorded with a director and an engineer in the room rather than a file handed over unheard. Synthetic voice is offered separately and is not counted anywhere on this page.