VoiceoverEdit and finish
AI voiceover generator that times narration to every scene
Muvomi's AI voiceover generator writes the narration for a film, records it in one of 20 English voices and lays each line on its scene. Ask for another voice or a new read and it is recorded again.
What Voiceover can do
Narration written for the film
The director writes one connected read and lays it over the film in passages. A passage begins on the scene where its idea appears and can keep playing across the cuts that follow.
20 English voices
Fifteen American and five British voices, each described by its delivery: calm, warm, energetic, deep. A lifelike voice reads unless you choose another. Ask for one by name or describe the voice you want.
One continuous take
Twelve voices, the four Lifelike and the eight Natural read, perform up to nine lines as a single take, so pace and tone carry from one line to the next. The other eight record line by line.
Lines fitted to their scenes
A line is written for 2.3 words a second and starts about half a second after the cut into its scene. If the recording runs long, its pauses are closed up, then it is tightened by up to 10% at the same pitch, then read again slightly faster.
Speech is never cut off
A line that still does not fit is not clipped or pushed into the next scene. You are told which scene is short and by how many seconds, so you can lengthen the scene or shorten the line.
Your script, word for word
Paste a script with scene headings and narration lines, and those lines are recorded exactly as written, each inside its own scene.
Direct the read in chat
Ask for a warmer read, a shorter line in scene three, another voice or no narration. The script changes and the voice is recorded again; a recording made for an earlier script is never played over the new one.
An even level over the music
Every line is set to the same loudness, with its loudest syllables held under a ceiling. The film's music drops by up to 76% while a line is spoken and comes back in the gaps.
Templates cut to the voice
28 motion templates carry a voice-over that reads what each beat shows. There the film follows the voice: a line starts with its beat, and the beat holds until the line has been said.
How Voiceover works
Describe the film
Open the AI Product Director and give it a brief, a script or a web address, and a length from 12 seconds to 5 minutes.
The narration is written and recorded
The film is planned scene by scene, its narration is written, and the voice is recorded and laid over the scenes.
Listen, then direct
Play the film. In chat, ask for a different voice, a rewritten read or a change to one line, and listen again.
Export with or without the voice
The export dialog has a switch for voice-over and one for background music. Render the MP4 with the sound you chose.
Voiceover formats and limits
Last updated .
| Detail | Voiceover |
|---|---|
| What goes in | A brief, a script or a web address, for a film of 12 seconds to 5 minutes |
| What comes out | A narrated film as MP4. The voice is one 24 kHz mono recording the length of the film, mixed with the music |
| Voices | 20, English only: 15 American, 5 British. 12 read in one continuous take and 8 line by line |
| A scene's line | Up to 300 characters, written for 2.3 words a second |
| Placement | A written line starts 0.5 seconds after the cut into its scene and is written to end 0.5 seconds before the next scene that speaks |
| Fitting a long line | Pauses closed to 0.3 seconds, then up to 10% faster at the same pitch, then a retake at 1.1 or 1.15 times the speed |
| Film | Up to 60 scenes of 1 to 60 seconds each |
| Music under the voice | Three instrumental tracks at a level of 0 to 65%, lowered by up to 76% under speech |
| Voiced templates | 28. A line starts 0.15 seconds after the cut into its beat, with at least 0.22 seconds between lines |
Who Voiceover is for
Product teams making an explainer
Turn a feature brief into a narrated walkthrough without booking a voice actor or a recording room.
Founders with a finished script
Paste the script with its scenes and hear it read as written, each line inside its own scene.
Marketers choosing a tone
Hear the same film in a calm American voice and a warm British one before deciding.
Agencies sending a first cut
Send a client a film with a voice on it, then export the music-only version if they are bringing their own narrator.
People working from a template
Fill in a voiced template's words and record it; the film retimes itself to the reading.
AI voiceover generator questions, answered
What does the AI voiceover generator do?
It gives a Muvomi film its narration. A script is written for the film, or yours is used, one of 20 voices reads it, and each line is placed on its scene and mixed over the music.
Which voices and accents can I choose?
Twenty voices, and all of them speak English. Fifteen have American accents and five are British: Colin, Gemma, Draco, George and Alice. No other language is available.
Can I use my own voice or clone a voice?
No. You cannot upload a voice sample or copy a person's voice. To use your own recording, import it into the video editor as an audio clip and cut the picture to it.
How is the narration timed to the scenes?
Each line belongs to the scene it starts on and begins about half a second after the cut. It may run on over silent scenes, and it is written to finish before the next scene that speaks. Preview and export play the same single recording, so both place every line at the same moment.
What happens if a line is too long for its scene?
First its pauses are shortened, then it is tightened by up to 10% without changing pitch, then it is read again a little faster. If it still does not fit, nothing is cut: you are asked to extend the scene or shorten the line.
Can I change the voice after the film is made?
Yes. Ask the director for another voice and the narration is recorded again for the film as it stands. You can also switch narration off and keep the music.
Does a voiceover cost extra?
A film's own voice comes with the film. Recording a voiceover, or a film's voice again, uses 20 credits for every minute of speech started, and the same words in the same voice again cost nothing.
What does the voiceover not do?
It is not a text-to-speech tool that hands back an audio file: the voice is recorded for a film and exported inside its MP4. You cannot set pitch, speed or emphasis by hand, and the video editor does not add an AI voice to footage of your own.
Voiceover
Give your film a voice.
Free to start, with 200 credits a month. Paid plans from $10 a month.