Google Vids voiceovers now sound more natural, with no setting to change

Google has moved the AI voiceovers in Google Vids onto a new Gemini speech model. It's live now for most paid Workspace plans and turns on automatically.

Google has upgraded the text-to-speech engine behind AI voiceovers in Google Vids, its video tool inside Workspace. Vids now uses a model called Gemini 3.8 Flash Lite TTS for voiceovers and for the narration on its AI avatars. Google says the new voices have more natural pacing and put the emphasis in the right places, so they sound less robotic.

The update is live now for both Rapid Release and Scheduled Release domains. It turns itself on: admins don't need to do anything and there's no setting for users. Google also says the audio generates faster, so you can change a line of the script and hear the new version almost straight away.

It covers Business Starter, Standard and Plus, the Enterprise plans, Education Plus and Nonprofits, as well as Google AI Pro and Ultra and several other editions. For a midsize business, this makes a scripted Vids video a more realistic option for training clips and internal updates. Those are jobs where booking someone to record a voiceover is often the reason the video never gets made.

Listen to the whole voiceover before you share it. Better pacing doesn't mean every word is right. Names, product terms and figures are where synthetic voices still slip. If a video goes to customers, check whether your industry expects you to say the voice is AI generated, and say so if you're unsure.

If you've written a script and never got around to recording it, open Vids and try the new voice on one short internal video first.