Google adds custom voices to its new speech models
Google began rolling out Gemini 3.8 Flash TTS and Flash-Lite TTS on September 23. The models support directed speech and two-speaker scenes.
Verified 12:25 AM PDT · 1 original sources
Google says voice replication requires a matching consent recording. Its AI Studio replication service excludes several regions, including the UK and EEA.
The launch includes company-reported quality comparisons. They do not establish how every customer script will sound. Access varies by product and region.
Creators should test their own scripts for pronunciation errors and unintended speaker changes before replacing a production voice.
Audit the story
Original sources
Company claims remain company claims. Follow the reporting and judge the evidence directly.
Continue the edition