logo
All guides

AI voiceover for faceless videos: choosing a voice and a language

In a faceless video the voice is the only personality the viewer meets. An AI voice-over lets you narrate without recording yourself, but it still needs good choices: the language, the voice, the wording and the pacing. This guide explains those choices, names the languages listed in the Nullface code and shows how to add and check a voice-over in the Story Creator before you export anything.

Why the voice matters in a faceless video

Without a face on screen, the narration carries most of the tone of the video. A calm, measured voice suits history and explainers, while a quicker, lighter voice suits lists and quotes. Viewers rarely analyse this, but they notice when voice and subject clash. Choosing the voice is therefore part of defining the format of the channel, not a last technical step before export.

A synthetic voice has practical advantages. You can change a line without recording again, keep the same voice across many videos and produce narration at any hour. It also has limits: it cannot improvise, and a badly phrased sentence still sounds badly phrased. Plan to listen to every scene, and treat the voice as a tool that follows your script faithfully, whatever the script says.

Think about how the video will be watched, too. Some viewers may have the sound low or off, so narration that depends on perfect attention is fragile. Say the key point in a short sentence, support it with text on screen and keep the picture consistent with the words. Voice, text and image should agree, so that missing one of them does not lose the message of the scene.

Languages in the Nullface code

The voice-language list in the Nullface code names 30 languages: Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Filipino, Finnish, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tagalog, Tamil, Turkish and Ukrainian. This guide mentions no other language, because the code lists are the only source we use for it. Which voices appear for a language depends on the voice library at the time, so open the voice selector in the Story Creator and look before you build a channel around one language.

Writing in a language other than English works the same way as in English: write the script in the language of the video and choose a voice that matches. Keep sentences short and avoid wordplay that depends on spelling, which a voice reads flat. If you publish in more than one language, treat each as its own channel with its own format checklist, because idioms and pacing rarely transfer without changes.

Write for the ear, not for the page

Text written to be read often sounds stiff when spoken. Use short sentences, plain words and one idea per sentence. Avoid long lists inside one sentence, brackets and abbreviations that a voice may read in an odd way. Spell out numbers and names that are hard to pronounce if the voice stumbles on them. Read each line aloud yourself first: if you run out of breath, the voice will sound rushed.

Punctuation steers the rhythm. A full stop gives a pause, a comma gives a lighter one, and a very long sentence gives none. Use that on purpose, especially at the hook and before the key point of a scene. The script length calculator on this site turns your word count into a spoken duration at a pace you set, which helps you check that the narration fits the scenes before you generate it.

Add and check the voice-over in the Story Creator

In the story settings you choose the language, and the list of voices narrows to the ones that match. You can play a preview of a voice before you choose it. The voice is prepared for the preview without credits, and regenerating a voice costs 0.1 credits, so you can test freely and fix single lines. Here is a routine that keeps the checks short.

  1. Write the script in the language the video should be spoken in.
  2. Open the voice settings of the story and pick the language, so that the list shows matching voices.
  3. Play the previews of a few voices and choose the one that fits the tone of your niche.
  4. Generate the voice for the scenes and listen to every scene from start to finish.
  5. Rewrite and regenerate only the lines that sound wrong, instead of redoing the whole video.
  6. Export the finished video, which needs a subscription, once the narration sounds right.

Keep the voice consistent across a series

Once you find a voice that suits the channel, keep it. A recognisable voice gives a faceless channel a sense of identity that a changing voice would break. Note the language and the voice you used in your format checklist, so that you or a collaborator can reproduce it later. Change the voice only as a deliberate test, and give the test a few videos before you judge it.

Be honest with yourself about disclosure. Some viewers care whether narration is synthetic, and platforms may have their own rules about it. We hold no sourced statement on those rules, so we do not summarise them. Read the current rules of each platform at the source before you publish, and decide how you want to describe your narration in the video text.

FAQ

Common questions

Which languages are in the voice-language list?
The voice-language list in the Nullface code names 30 languages: Arabic, Bulgarian, Chinese, Croatian, Czech, Danish, Dutch, English, Filipino, Finnish, French, German, Greek, Hindi, Indonesian, Italian, Japanese, Korean, Malay, Polish, Portuguese, Romanian, Russian, Slovak, Spanish, Swedish, Tagalog, Tamil, Turkish and Ukrainian. Which voices appear for a language depends on the voice library, so check the voice selector in the Story Creator before you plan a channel around one language.
Do I need credits for the voice-over?
The voice is prepared for the preview without credits, and regenerating a voice costs 0.1 credits. Exporting the video needs a subscription. AI clips and AI music use credits. Check the pricing page for the current plans and the credits they include before you plan a series.
Will viewers notice that the voice is synthetic?
Some will and some will not. Choose a voice that matches the tone of your niche, read your script aloud first to fix awkward phrasing and listen to every scene before you export. Read the platform rules on disclosure at the source, because they can change.
Can I try the voice-over before I subscribe?
topic analysis, script and storyboard, clips from the public library and the editor preview work after sign-up without a subscription, and the voice is prepared for the preview without credits. Exporting the video needs a subscription. AI clips and AI music use credits. That lets you judge the narration in the editor preview before you decide whether to subscribe.
Can I change the voice later?
You can pick another voice in the story settings and generate the narration again, and regenerating a voice costs 0.1 credits. For a channel, though, a stable voice helps viewers recognise your videos, so treat a change as a deliberate test and give it several videos before you decide.

Newsletter

New video ideas and Nullface news by email

Leave your address and we will send you new video ideas and product news. Not signed up with Nullface yet? You do not need an account for this.

By submitting your address you agree that Nullface AI emails you new video ideas and product news. We first send one confirmation email, and you only receive the newsletter after you confirm. You can unsubscribe at any time with the link in every email. How we handle your data: Privacy Policy.

Make your first video

Open the Story Creator and turn the topic into a storyboard. Create your account first; the button takes you to the Story Creator right after sign-up.