Our model faithfully reproduces original voices in our voice clones, emulating everything from tone, speed, inflections, accent, breathing patterns, mouth clicks, and noises in the surrounding environment. With our instant cloning capability, all you need is 5 seconds of source audio and you’ll have a clone ready in seconds!
We recently retired the ability to create professional voice clones. Instant clones offer better speaker similarity while requiring significantly less source audio.View more details about this update in our changelog
Below, we provide some tips for optimizing input quality, which, in turn, will shape output quality.
Does your clone not sound like the source audio at all? Make sure your source audio does not contain any segments with music, other people talking, or prolonged silence.
  • Audio Quality
    • Use clear audio with minimal background noise
    • Record in an acoustically-treated room or use exterior padding (e.g., blankets) to reduce echoes and background noises
    • Position yourself 6-12 inches away from the microphone
    • Consider using premium recording equipment:
      • XLR microphone (e.g., Shure)
      • High-quality audio interface (e.g., Focusrite)
      • Pop-filter to eliminate unwanted sounds
    • Note: Your computer’s built-in microphone will also work well!
  • Recording Technique
    • Speak at the same speed you want the voice clone to use
    • Use a wide range of emotions, especially those you want the voice clone to adopt
    • Maintain the accent you want the AI to reproduce
    • If you’re unhappy with the results, you can always record again:
      • Adjust your speaking style (e.g., if the clone sounds too high-pitched, try speaking with a deeper intonation)
      • Experiment with different emotional ranges
      • Fine-tune your accent or speech patterns