Voice Over for TikTok, Reels and Shorts

A short video voice over is an audio narration laid over a vertical video lasting a few seconds to a few minutes, which explains, tells a story or accompanies the images.

It replaces the on-camera presenter or complements it. Today it can be recorded by a person or generated by artificial intelligence from a text, which lets you produce videos faster and with a consistent voice from one post to the next.

1. Write a short script that hooks from the first sentence

In a short video, every word counts. Count about 150 words per minute when spoken, or roughly 40 words for 15 seconds.

• Open with the promise or the question, not with an introduction.

• One idea per video, one idea per sentence.

• Short sentences, spoken vocabulary.

• Finish with a simple action: follow, try it, read more.

Tip: read your script aloud while timing yourself. If it runs over the planned length, cut.

2. Choose the voice and the tone

An energetic voice suits dynamic content, a calmer one suits explanations or tutorials. Test several voices on the first sentence of your script before deciding. If you publish regularly, keep the same voice so your audience recognizes you.

3. Generate the voice over

Paste your text into the tool, select the voice and start the generation. The result is ready in moments.

4. Listen, adjust, regenerate

Listen to the whole audio, preferably with headphones, then fix whatever stands out:

• a mispronounced word: rephrase it or write it phonetically,

• a misplaced pause: add or remove a comma or a period,

• a flat passage: split the sentence or vary the punctuation.

Two or three iterations are usually enough.

5. Edit the video with captions and export in vertical format

Place the voice over on your visuals, add on-screen captions, then export in 9:16 vertical format. Many people watch without sound: captions keep your message readable. With Incarneo, you can also pair the voice with an animated avatar to get a talking video directly.

• Hook in under 3 seconds: the first sentence states the topic or asks a question.

• Keep a brisk pace: cut unnecessary silences, but leave a breath between two ideas.

• Sync voice and visuals: change shots when a new idea begins.

• Add captions: many videos are watched without sound.

• Keep captions out of masked areas: each platform's buttons and caption cover the bottom and edges of the screen.

• Lower the music under the voice: the voice must stay clear, even on a phone.

Count about 150 words per minute when spoken:

• 15 seconds: about 40 words, ideal for a hook or a teaser.

• 30 seconds: about 75 words, ideal for a tip or a quick demo.

• 45 seconds: about 110 words, ideal for a mini-tutorial.

• 60 seconds: about 150 words, ideal for an explanation or a story.

Maximum lengths vary by platform and change regularly: check them in each app before publishing.

• Reusing text written for an article without shortening it or making it more conversational.

• Starting with an introduction ("Hi, today I'm going to talk about...") instead of the main idea.

• Choosing a voice without testing it on a real sentence from your script.

• Forgetting captions or placing them where the interface hides them.

• Cloning a voice without the person's consent: always ask for their agreement.

• Not flagging an AI-generated voice when the platform or context requires it: some platforms offer labeling for AI-generated content.

Incarneo is an AI studio that lets you clone a voice and animate an avatar to create voice overs and talking videos in a few minutes, alone or in a team.

For your TikTok, Reels and Shorts videos, you can:

• turn your script into a voice over,

• keep the same voice from one post to the next thanks to the cloned voice,

• pair the voice with an avatar, created from a front-facing photo, to get a talking video,

• reuse your avatars and voices for your next content,

• export the audio or video and publish it on your platforms.