We’re making some improvements to our AI voice system to deliver faster, more natural-sounding results and a better overall experience when working with AI speakers. As part of that effort, we’re retiring some older AI speakers created with our original voice model.

Any AI speakers created with the older version of our voice clone model will be retired and no longer available after June 30, 2025. If you’re seeing this message, it means there's an AI speaker on your Drive created with the legacy model–these speakers won’t work in new projects after that date.

You’ll still be able to listen to and export any content you’ve already created with this voice—but you won’t be able to generate new audio with it going forward. Your training data and project will also be removed after June 30.

What to do next

If you’d like to continue using AI voices in your projects, you’ll need to recreate this speaker using our latest model. It’s quick to set up, and you’ll get improved pacing and more natural-sounding results.

We have a step-by-step process for available in our Creating a new AI Speaker Help Center guide.