Custom Voice

Mirako allows you to create custom voice profiles that can be tailored to your specific needs. These voices can be created by cloning existing voices or by uploading your own audio samples.

Learn more or get started with Voice Cloning.

List Your Custom Voices

You can get a list of your custom voice profiles using Mirako CLI:

sh
mirako voice list

Or, you can integrate the functionality into your application using the Mirako API.

python
def get_custom_voices():
    """Get list of your custom voice profiles"""
    
    response = requests.get(
        f"{BASE_URL}/v1/voice/profiles",
        headers=headers
    )
    
    if response.status_code == 200:
        voices = response.json()['data']
        
        if voices:
            print("🎙️ Your Custom Voices:")
            for voice in voices:
                print(f"  • {voice['name']} (ID: {voice['id']})")
                print(f"    Status: {voice.get('status', 'Unknown')}")
                if voice.get('created_at'):
                    print(f"    Created: {voice['created_at']}")
        else:
            print("No custom voices found")
        
        return voices
    else:
        print(f"Error: {response.text}")
        return []

# Get your custom voices
custom_voices = get_custom_voices()

Delete Custom Voice

To delete a custom voice, delete it by using the following Mirako CLI command:

sh
mirako voice delete <voice_id>

or use the following API request:

python
def delete_custom_voice(voice_id):
    """Delete a custom voice profile"""
    
    response = requests.delete(
        f"{BASE_URL}/v1/voice/profiles/{voice_id}",
        headers=headers
    )
    
    if response.status_code == 204:
        print(f"Custom voice with ID {voice_id} deleted successfully.")
    else:
        print(f"Error deleting custom voice: {response.text}")
# Delete a custom voice by ID
# CAUTION: This action is irreversible.
delete_custom_voice("your_custom_voice_id_here")

Using Custom Voices

Like premade voices, you can use custom voices by referring to their voice_id in your API requests for text-to-speech or interactive sessions.

Dive Deeper