Skip to content

πŸ€– Explain with AI

Voice Cloning

Create custom voices tailored to your brand. Clone any voice for use across all your AI agents.

Overview

Voice Cloning provides:

  • Custom Voice Creation β€” Clone any voice via browser recording
  • Multi-Language Support β€” English, Spanish, Portuguese, French
  • Voice Management β€” Organize and manage cloned voices
  • Agent Integration β€” Use cloned voices across all voice agents
  • Workspace Sharing β€” Share voices across team members

Built-in Voices

Five professionally crafted voices are available out of the box:

Voice Character Best For
Eve Warm, professional Receptionists, customer support
Ara Expressive, versatile Sales, marketing
Leo Confident, articulate Collections, operations
Rex Deep, authoritative Compliance, legal notifications
Sal Friendly, upbeat Appointments, reminders

Creating Custom Voices

Recording Your Voice

  1. Navigate to Voice Cloning in the AI Agents section
  2. Click Record Your Voice
  3. Read the provided script (or record your own content)
  4. Click Stop when finished
  5. Play Preview to review your recording
  6. Click Re-record if needed

Voice Configuration

Field Description
Voice Name Unique name for your voice (e.g., “My Support Voice”)
Language English, Spanish, Portuguese, or French
Audio Source Browser recording or uploaded file

Cloning Process

  1. Click Clone Voice
  2. The audio is processed and uploaded
  3. A unique voice ID is generated
  4. The voice becomes available for use across all agents

Supported Audio Formats

Format Supported
WebM Yes (browser recording)
WAV Yes (upload)
MP3 Yes (upload)

Managing Cloned Voices

Voice List

The voice list displays all cloned voices:

Column Description
Name Voice name
Language Voice language
Status Active, Processing, or Deleted
Created Creation date
Voice ID Unique identifier

Voice Actions

Action Description
Copy Voice ID Copy the voice ID for manual configuration
Delete Remove the voice (soft delete)

Deleting Voices

Deleting a voice sets its status to “deleted” rather than removing it from the database. This prevents orphaned references in existing agent configurations.

Using Cloned Voices

Assigning to an Agent

  1. Navigate to AI Agents β†’ My Agents
  2. Select or create an agent
  3. Go to Channel β†’ Phone
  4. In the Voice dropdown, select your cloned voice
  5. Save the agent configuration

Voice Selection

When configuring a voice agent, you can choose from:

  • Built-in Voices β€” Eve, Ara, Leo, Rex, Sal
  • Cloned Voices β€” All voices you’ve created in your workspace

Workspace Sharing

Cloned voices are workspace-scoped. All team members in the same workspace can use voices created by any member.

Technical Details

Voice Processing

Step Description
Audio Upload Recording is converted to WAV format
Voice Analysis AI analyzes vocal characteristics
Voice ID Generation Unique identifier created
Status Update Voice marked as “active”

Voice Quality

For best results:

  • Record in a quiet environment
  • Speak clearly and at a natural pace
  • Use a high-quality microphone
  • Record at least 30 seconds of speech
  • Avoid background music or noise

Supported Languages

Language Code Quality
English en Excellent
Spanish es Excellent
Portuguese pt Good
French fr Good

API Reference

Get Voice ID

To use a cloned voice programmatically, copy the Voice ID from the voice list and use it in agent configuration:

voice_id: "your-voice-id-here"

Voice Listing

Cloned voices are listed via the Supabase client, scoped to the user’s workspace(s) via team member lookup.

Best Practices

  1. Test before deploying β€” Use the playground to verify voice quality
  2. Match language to audience β€” Use the correct language code for your target market
  3. Consistent recording β€” Use the same microphone and environment for all recordings
  4. Descriptive names β€” Name voices clearly (e.g., “Customer Support Voice - EN”)
  5. Monitor quality β€” Regularly test cloned voices for degradation

Troubleshooting

Voice Not Appearing

  • Verify the voice status is “active”
  • Check you’re in the correct workspace
  • Refresh the voice list

Poor Voice Quality

  • Re-record in a quieter environment
  • Use a higher-quality microphone
  • Record longer samples (30+ seconds)
  • Ensure the language setting matches the recording

Voice ID Not Working

  • Verify the voice ID is copied correctly
  • Check the voice hasn’t been deleted
  • Ensure the agent is configured to use the correct voice provider

What’s Next


Still questions? Ask on Discord or explore tutorials