Skip to content

VOICE CLONING / CREATOR GUIDE

How to clone a voice for AI dialogue and video dubbing

AI voice cloning turns an authorized reference recording into a reusable cloned voice. In VoiceCast, bring that voice into character dialogue, audio dramas, narration and subtitle-based video dubbing.

Updated October 27, 2026

Create a cloned voice ↗

From a reference recording to a finished scene

  1. Prepare an authorized voice recording

    Use one clear speaker, with little background noise or music. Choose a 6–60 second recording up to 10 MB. Use your own voice or obtain the speaker’s explicit permission. A clean, natural sample gives you a better starting point.

  2. Create and name your cloned voice

    Sign in with a verified email, open Voice cloning, enter a voice name and choose a generation mode. Upload a recording or record directly in the browser, confirm your permission and create the voice. Browser recording requires microphone access.

  3. Choose the right generation mode

    English · Fast is intended for English speech and does not support a per-line emotion override. Multilingual · Expressive supports delivery guidance. Review your chosen language and preview the result; style guidance does not guarantee a particular performance.

  4. Preview before making a full scene

    Find your saved voice under My voices and click Preview. Creating preview audio uses credits; a cached preview can be replayed without generating it again. Review pronunciation, clarity and likeness before producing a longer scene.

  5. Cast the voice in dialogue or video

    In Theater, parse a script, open the voice selector and choose your cloned voice. In the voice library, My voices shows the current account’s ready cloned voices. For video dubbing, upload a video and recognize or import SRT subtitles, then expand a line in the Subtitle dubbing list to select the cloned voice.

  6. Direct the scene and export

    Edit dialogue, pacing and pauses; adjust emotion when the selected mode supports it. Regenerate changed lines to hear new settings. Theater exports MP3, WAV or FLAC audio and SRT subtitles. Video dubbing exports MP4, MOV, MKV or WebM video and WAV, MP3 or FLAC dubbing audio. Save the project to return to its text and settings.

Voice cloning FAQ

Can I save more than one cloned voice?

Yes. The current implementation supports up to 12 saved cloned voices per account. Each voice keeps its own reference recording. Available capacity is shown on the creation page.

Are my cloned voices public?

Cloned voices are associated with the signed-in account. Other accounts cannot select them from the public voice library. Reference recordings are stored on the application server for generation; they are not published as tutorial samples.

Is voice cloning free?

Creating and saving a reference voice is separate from generating speech. Speech generation uses the usual VoiceCast credits. Check the estimate, plan rules and credit packs before generating. Current checkout is a test workflow.

Does cloning translate or lip-sync a video?

No. Video dubbing reads the subtitles you provide or edit. Prepare translated dialogue yourself. Automatic translation and lip synchronization are not included.

Can I replace or delete a reference recording?

Yes. Open Voice cloning to replace the reference, rename or delete a cloned voice. Replacing the reference changes future generations; deleting a voice prevents projects from generating it again. Existing downloaded audio is not removed.