Introducing Song Creator Pro — create music with AI, locally on your device. Try it now →

Voice Assignment

Assign voices to speakers, characters, and sections in your project for multi-narrator audiobooks and scripted content.

Voice assignment lets you map different voices to different parts of your project. Use any combination of built-in voices, cloned voices, and designed voices from the Lab.

Voice Sources

You can assign voices from any of these sources:

  • Built-in voices from the TTS tab in Lab
  • Cloned voices from the Clone tab in Lab
  • Designed voices from the Design tab in Lab
  • Community voices imported from Voice Search

Mix and match freely. For example, use a cloned narrator voice for prose and designed character voices for dialogue.

Assigning Voices

Voice assignment works at three levels:

Default Voice

Set a default voice for the entire project. This voice is used for all segments across all sections unless you override it. For single-narrator content, this may be the only voice you need.

Per-Segment Overrides

Override the default voice for any individual segment. Select the segment and choose a different voice from the voice dropdown. This is useful for segments that belong entirely to a different speaker or character.

Inline Voice Spans

For the most granular control, highlight a span of text within a segment and assign a different voice to just that selection. This lets you handle cases like:

  • Quoted dialogue embedded in narration
  • A character interrupting the narrator
  • Multiple speakers within a single paragraph

The surrounding text keeps its current voice while the highlighted span uses the override voice.

Workflow Tips

  • Save voices to your library in the Lab before starting a project so they are ready to assign
  • Preview before bulk generation. Generate a few key segments first to make sure each voice sounds right in context before generating the full project