1. Stability (Balance Between Consistency and Variation)
Stability controls how consistent the voice sounds.
- High stability → smooth but robotic
- Low stability → expressive but sometimes unstable
Best Range:
👉 30% – 60%
Lowering stability introduces small imperfections exactly what makes speech sound human.
If you're working on content like storytelling or narration (as discussed in AI Voices for Storytelling), slightly lower stability helps create emotional depth.