Skip to content
Inworld AI· inworld·· 12 d agoPickAI score38

Inworld breeds synthetic voices by having an LLM iteratively rewrite descriptions, picking 12 after 26 rounds

Our Head of Research, @altsoph, shared an experiment in “breeding” synthetic voices to bring more variety to our TTS library.

AI summary

Inworld head of research Aleksey used Inworld Voice Design to generate voices from text descriptions, then had an LLM rewrite and combine those descriptions based on evaluation results. After 26 rounds of iteration, 214 voices passed automated checks, and human review picked 12 of them for the voice library. Each round tests audio quality and how much a voice differs from those already in the library; the full pipeline and sample clips are published on his Substack.

Why it matters

The original lays out the full pipeline of using an LLM to iteratively rewrite descriptions and filter voices on two fronts, a useful reference for building your own voice library.

Source: Inworld AI · x.com