Voice Customization

Voice Customization lets you create custom variants of a voice – adjusting its speed and stability – without changing the voice itself. Use it to make the same voice sound calm and steady for a training course, or more dynamic for a promo video.

Where to find it

1. Open your video editor and click to select a voice.

2. In the select a voice window, look for the slider icon  under a voice's name – it means this voice supports customization.

3. Once you've saved a variant, a second icon appears (microphone icon with a number) showing how many custom variants that voice already has – hover it to see "This voice has custom variants".
   

4. Click the voice to open the variant panel at the bottom.

Each variant row shows three controls next to its name:

• ⓘ  — variant details

• ✏️ — rename the variant 

•  — remove the variant

Creating a variant 

Each voice can have up to 3 variants (plus the Original). For each variant you can adjust:

Speed – controls the speed of the generated speech. Values below 1.0 slow the speech down, values above 1.0 speed it up. Extreme values may affect the quality of the generated speech.

Stability – controls how consistent the voice sounds between re-generations.

• Higher stability = more consistent between re-generations, but can sound a bit monotone

• Lower stability = more natural variation

Tip: on longer text fragments, we recommend lowering the Stability value – high stability tends to sound monotone on longer speech.

Steps

1. Click Add Variant to create a new one – up to 3 per voice

2. Adjust the Speed and Stability sliders

3. Click ▶ Play to preview the result

4. Click ↻ to reset the variant to previous if you want to start over

5. Click Save variant to keep it, or check Apply to all slides and click Apply voice to use it across your whole video

Tips

• Preview before applying. Small slider changes can noticeably change the result – always listen with Play before saving.

• You get 3 variants per voice. Use them to compare a few options side by side (e.g. one calm, one energetic) before deciding which fits your video.

• Apply to all slides applies the variant to every slide using that voice – useful for consistency, but turn it off if you want different variants on different slides. 

• Same settings can still sound slightly different between renders, especially at lower Stability – this is expected behavior of AI voices, not a bug.