Voice cloning

Record a few seconds of clear speech or import clips, and Avocado turns them into a voice you can use with every model that supports cloning. Only clone voices you have permission to use.

Voice design

No recording needed. Describe the voice you want, like "a calm narrator in her forties with a light Irish accent", audition it, and save it to use again.

Train your own voice

Cloning copies a voice from a short clip. Training goes further: give Avocado a few minutes of your speech and it teaches the model how you sound and how you speak, right on your Mac.

Private on-device TTS

Avocado makes speech on your Mac, not on a server. Your scripts, recordings, voices and takes stay in Avocado, and there is no account to create.

Dialogue

Write a script with turns for each speaker, cast a voice for each one, and Avocado generates the whole conversation in one take. Good for podcasts, explainers and audio drama.

Edit takes

Got the words right but the mood wrong? With Step Audio EditX, open any take up to 30 seconds long, change its emotion, style or pace, and keep the original beside it.

Sound effects

Type what you want to hear, choose a length up to 60 seconds, and Avocado generates it on your Mac with MOSS Sound Effect. Your effects sit in the same projects as your voice takes.

Avocado is almost here

Free for Apple silicon Macs running macOS 15 or later. Coming soon.

Coming soon for Mac