Use case · YouTube & video
Write the script, pick a voice that fits your channel, and export narration paced to your cut. When the video's done, dub it, so the same channel speaks Spanish, Hindi and Portuguese.
For YouTubers, faceless channels, explainer studios, agencies
Three years ago, this valley was underwater.
Today it feeds a city of two million people.
We didn't drain it. We taught it where to go.
01What changes
02How it goes
03What it costs
Starter makes about 53 a month. A standard voice is a credit a word; premium voices 1.5.
04What you'll use
05Questions
An AI voiceover is only as good as the script and the direction behind it. This guide walks through writing for the ear, choosing a voice that suits your channel, timing narration to an edit, and taking a finished video into other languages.
Viewers can't re-read a sentence they missed, so narration works best in short sentences with one idea each. Put the important word at the end of the sentence, where the voice naturally lands, and read your script aloud once before you generate anything: anything you stumble on, a voice will too.
A useful rule of thumb is 150 words a minute. A 10-minute video is around 1,500 words, an 8-minute explainer around 1,200. If your script runs long, cut it rather than speeding the voice up: fast narration tires people out.
Documentary and history channels tend to suit warm, lower, unhurried voices. Tech reviews and tutorials do better with clear, mid-paced, conversational ones. Kids' content and shorts can carry more energy. Listen to several candidates reading the same paragraph of your script, not the voice's own sample, before you decide.
Once you've chosen, save it as your default voice in Settings so every new project starts with it. Consistency is what makes a faceless channel recognisable. If you'd rather it were your voice, you can clone it from a minute or two of clear speech.
In the Studio every paragraph is its own line, so you can set the pace of each one, add a pause where the picture needs to breathe, and emphasise the word that matters. Because lines are generated separately, fixing one never shifts the others: change a sentence, regenerate that line, and the rest of the timeline stays put.
Sound belongs in the same place. Describe an effect, such as a whoosh, a page turn or city ambience, and it's generated on its own lane with a level you can hear against the voice. Export one mixed file, or stems per voice if your editor prefers to mix.
Once the video is finished, dubbing translates what's said and re-voices it in a voice made from the original, so the channel sounds like itself in Spanish, Hindi or Portuguese. Upload the video or paste a link, choose up to three languages at a time, and open any dub in the Studio to fix a name or a phrase.
Dubbing costs 5,000 credits a minute for each language, charged by the second, and failed dubs are refunded automatically.
Audio you make on any paid plan can be used commercially, which includes monetised YouTube videos, sponsored content and client work. Audio made on the free plan is for personal use, so it's the right place to find your voice and workflow before you publish.
06More use cases
1,000 free credits to try everything. Sign up and your first youtube & video project is ready to write.
Start free: 1,000 credits