Why Most of the World Was Never Dubbed

01.The cost of one language

Full human dubbing, translation plus voice, runs $70 to $150 per finished minute. For one 60-minute video, that is $4,200 to $9,000, for one target language. Ten languages multiply it by ten.

A blockbuster can carry that for its biggest markets. A documentary, a talk show or a season of vertical drama usually cannot carry it for even one.

02.The people behind every language

A dub is a production. Each language needs its own translator and adapter, a casting call, voice actors for every role, a director, a studio and a mixer. Catherine Retat, Netflix's senior director for global content localisation, put the scale plainly at a localisation roundtable in Seoul: it usually takes 50 to 60 people to produce a single dubbed version, so dubbing into ten languages means about 500 people, according to Digital Today.

That craft is real, and it is why the best dubs sound the way they do. It also means every new language is a new team to find, book and direct.

03.The time

Each language runs its own schedule: adapt the script to the lip movements, cast, record line by line, direct the takes, mix against the original. Release dates rarely wait for the slowest language, so the list gets cut to the languages that can finish in time.

04.Who gets left out

Put cost, people and time together and the result is the map we have today:

• Top productions dub into a handful of their biggest markets.
• Smaller productions, shows and vertical dramasmostly don't dub at all; subtitles, at best.

The audience is not waiting in English. English accounts for 25.9% of internet users in Internet World Stats' estimate as of March 31, 2020, cited on Wikipedia: about 74% of internet users don't speak it. For most of them, most of the world's video has never been dubbed.

05.Machine learning for the tedious part.

Most of the work in a dub repeats: carrying every line into the new language, in the speaker's voice, with the lips moving to the new words. That is the part machine learning can take. Familiar translates the words, carries the speaker's own voice from the clip itself, and syncs the lips to the new words, into 30 languages. The actor, the performance, the timing and the scene sound stay as they were.

The team stays in charge. The dub pauses at a review screen before it renders, so your creative directors read and edit the translation against the original, and a line that can land more than one way comes back as options to pick from.

The idea fits in one line: the same Familiar performance for every audience.

06.Where the data comes from

All of the data we trained with ourselves are actually generated by the Familiar team.

Nothing new is made up and there is no prompt: your video goes in, and the same performance comes out in another language. It is not generative AI in the traditional sense. It is translation, so a show made for one country can reach the next.

We're another tool for the creative process.

QUESTIONS

REFERENCES

1. [1]Digital Today, Netflix says it deploys up to 60 people per dubbed language, works with 122 partners worldwide, October 15, 2026
2. [2]Wikipedia, Languages used on the Internet (Internet World Stats, March 31, 2020), read October 24, 2026
3. [3]How Much Does Dubbing Cost?
4. [4]Why Familiar Is Not Traditional Generative AI