Voice Selection & Cloning

The voice of your assistant determines the personality of your brand. Caller AI offers a vast library of high-quality pre-made voices, as well as the ability to clone your own voice for a fully personalized experience.

Selecting a Standard Voice

Caller AI integrates with top-tier voice providers to give you access to hundreds of accents, ages, and tones. To select a voice:

  1. Navigate to your Assistant Settings.
  2. Click on the Voice tab.
  3. Use the filters to search by Language, Gender, or Accent (e.g., American, British, Australian).
  4. Click the "Play" icon to preview the voice before saving.

Instant Voice Cloning

You can clone your own voice or the voice of a team member to create a digital replica. This is ideal for founders, sales leaders, or specific brand personas.

How to Clone a Voice

Warning: Audio Quality Matters
Do not upload audio with background music, heavy echo, or other people talking. Dirty audio results in a "robotic" or "glitchy" clone.

Fine-Tuning Voice Settings

Once you have selected or cloned a voice, you can tweak its performance characteristics to better suit your use case.

Setting Description
Stability Determines how consistent the voice sounds.
High: Monotone, consistent.
Low: More expressive, but can be unpredictable.
Similarity Boost (Cloned voices only) How closely the AI tries to match the original audio sample.
Note: Setting this too high can introduce audio artifacts.
Speed Adjusts the speaking rate.
0.9x - 1.0x is usually best for support.
1.1x works well for high-energy sales.

Matching Voice to Use Case

For Cold Outbound: Use a casual, slightly faster voice with lower stability (more expression) to sound like a real person calling on a cell phone.

For Inbound Support: Use a clear, deep, or authoritative voice with higher stability to convey trust and professionalism.