Skip to main content
Clone voices to create unique, personalized text-to-speech that sounds like you, a team member, or a custom brand voice. Use these voices for consistent audio content across all your projects.
You don’t have to clone a voice to get started. A large library of ready-made voices is available to use immediately in Audio. Clone a voice only when you want a specific person’s or brand’s voice. Cloned voices are private to your company; ready-made voices are available to everyone.

​
What Is Voice Cloning?

Voice cloning uses AI to learn from audio samples and create a synthetic voice that sounds like the original speaker. Once created, you can use this voice to generate any text as speech.

​
Why Clone a Voice?

  • Brand consistency: One voice across all content
  • Personalization: Use your own voice without recording everything
  • Scalability: Generate hours of content from minutes of samples
  • Multiple languages: Some voices can speak languages the original speaker doesn’t

​
Creating a Voice Clone

​
Step 1: Open Voice Cloning

  1. Select your company from the sidebar
  2. Click Voices (under the Content Engine menu)
  3. Click Create Voice (or the + button)

​
Step 2: Prepare Your Audio Samples

You’ll need to provide audio samples for the AI to learn from. Requirements:
  • High-quality recordings
  • Clear speech without background noise
  • 1-5 minutes of audio total
  • Natural conversational speech

​
Step 3: Upload Your Samples

  1. Click Upload Audio
  2. Select your audio file(s)
  3. Supported formats: MP3, WAV, M4A

​
Step 4: Configure Settings

  • Voice Name: What you’ll call this voice
  • Description: Notes about the voice (optional)
  • Remove Background Noise: Enable if recordings have some noise

​
Step 5: Create the Voice

  1. Click Create Voice
  2. Wait for processing (can take several minutes)
  3. Your voice is ready when the status shows complete

​
Recording Great Voice Samples

The quality of your clone depends on the quality of your samples.

​
Recording Environment

​
Recording Equipment

Good options:
  • Professional microphone
  • Quality headset microphone
  • Modern smartphone in a quiet room
Avoid:
  • Built-in laptop microphones
  • Speakerphone recordings
  • Phone calls or video conference recordings

​
What to Say

Record natural, conversational speech:
  • Read a few paragraphs from a book or article
  • Tell a story about your work
  • Explain something you know well
  • Answer questions out loud
Include variety:
  • Statements and questions
  • Different emotions (neutral, happy, serious)
  • Various sentence lengths
Avoid:
  • Whispering or shouting
  • Extreme emotions
  • Heavy accents you don’t normally use
  • Reading in a monotone voice

​
Sample Script

If you need something to read, try variations like:

​
Testing Your Voice Clone

After creation, test before using in production:

​
Step 1: Generate Test Audio

  1. Find your voice in the list
  2. Click Test or the play button
  3. Enter test text:
  1. Click Generate and listen

​
Step 2: Evaluate the Result

Check for:
  • Does it sound like the original speaker?
  • Is pronunciation correct?
  • Does it sound natural, not robotic?
  • Are there any strange artifacts?

​
Step 3: Refine If Needed

If quality isn’t great:
  • Try uploading more sample audio
  • Ensure samples are high quality
  • Remove any samples with background noise
  • Re-create the voice with better samples

​
Using Your Custom Voice

Once your voice is created, use it anywhere you generate audio:

​
In the Audio Generator

  1. Go to Audio
  2. Click Generate Audio
  3. In the voice selection, find your custom voice
  4. Write your text and generate

​
Your voice works for:

  • All text-to-speech generation
  • Any length content
  • Multiple languages (if the model supports it)

​
Managing Your Voices

​
Viewing Your Voices

  1. Go to Voices
  2. See all your created voice clones
  3. Status shows if each is ready to use

​
Deleting a Voice

  1. Click the three-dot menu (⋮) on the voice
  2. Select Delete
  3. Confirm
Note: Deleting a voice is permanent. Audio already generated with it will still work.

​
Voice Quality Tips

​
1. More Samples = Better Quality

While 1 minute can work, 3-5 minutes of varied speech usually produces better results.

​
2. Consistent Recording Conditions

If recording multiple sessions:
  • Use the same microphone
  • Same distance from mic
  • Same room/environment
  • Similar time of day (voice changes throughout the day)

​
3. Natural Speech Patterns

Don’t over-enunciate or speak unnaturally. The AI learns from how you actually speak, so be natural.

​
4. Clean Audio Editing

If you edit your audio:
  • Remove long silences
  • Cut out coughs, “ums,” and mistakes
  • Don’t over-process with effects

​
Ethical Use Guidelines

​
DO use voice cloning for:

  • Creating your own voice clone
  • Voices of people who have given consent
  • Fictional or synthetic brand voices
  • Accessibility applications

​
DON’T use voice cloning for:

  • Impersonating others without consent
  • Creating deceptive content
  • Fraud or scams
  • Spreading misinformation
Important: Only clone voices of people who have given explicit permission.

​
Frequently Asked Questions

Q: How long does voice creation take? A: Usually 5-15 minutes for processing. Q: Can I update a voice with more samples? A: Currently, you’d need to create a new voice with the combined samples. Q: How realistic are voice clones? A: Quality varies based on your samples. Good samples produce very realistic results. Q: Can the cloned voice speak other languages? A: Some models support multiple languages, even if your samples are in one language. Q: Is there a limit to how many voices I can create? A: Check your subscription for limits on voice clones. Q: Can I share my voice clone with others? A: Voice clones are currently limited to your account/company.

​
Next Steps