Short Reference Recording
Record about 10–20 seconds of clear speech. This becomes the reference the installed synthesis engine uses for the voice.
Record a 10–20 second reference on an Apple Silicon Mac. Yaps saves that WAV and its metadata locally, then passes the reference to the installed synthesis engine when you choose the voice in a supported read-aloud or Studio workflow.
“Hey everyone, welcome back to another episode. Today we are talking about something I have been thinking about a lot lately.”
Voice Cloning is available in Yaps desktop on Apple Silicon (M1 or later). Check the app for current plan access. Yaps stores the reference recording locally and uses it with the installed voice engine at synthesis time.
A short reference recording, local storage, and reference-guided synthesis in supported Yaps voice workflows.
Record about 10–20 seconds of clear speech. This becomes the reference the installed synthesis engine uses for the voice.
Yaps does not train a new model for every voice. It uses the shared installed Chatterbox engine and supplies your saved reference at synthesis time.
The local engine conditions its output on the short reference recording when you select that voice.
Yaps stores the reference WAV and its metadata on your Mac rather than creating or uploading a separate personal model file.
Once the local synthesis engine is installed, reference-guided generation can run without an internet connection.
Choose the saved voice reference in supported Read Aloud and Studio voice pickers; it is not a system-wide voice or a separately trained model.
When the voice is yours, the content feels different. More honest. More you.
Generate audio segments in your own voice for podcast intros, ad reads, or episode previews. Your listeners hear you, even when you're not in the recording booth.
“Hey everyone, welcome back to another episode. Today we're talking about something I've been thinking about a lot lately.”
Authors can narrate their own books without spending weeks in a studio. Write the chapter, select your cloned voice, and export. Update a passage and regenerate in seconds.
Create voiceovers for tutorials, product walkthroughs, and presentations in your own voice. Your audience connects with a voice they recognize.
For users who may be losing their voice due to medical conditions, voice cloning preserves a digital version of how they sound. A deeply personal use of private, local technology.
Solo creators, educators, and small business owners can maintain a consistent voice across all their audio content without recording every piece manually.
Create voice messages, greetings, or presentations that sound like you. When a generic voice would feel cold, your own voice adds warmth and authenticity.

Yaps is now on iPhone and iPad. It is a keyboard, so it types into any app you already use, and the speech recognition runs on the device rather than on somebody else's server. Here is what shipped, what it costs, and what it still does not do.

Every AI note tool built for therapists asks for the same thing first: send us the session. Here is the other route. Dictate your progress note by voice, have it cleaned up on your own machine, and keep it in a plain-text vault that no vendor can read, because no vendor ever receives it.

Ministry produces two very different kinds of note. One is a sermon idea that arrives on a walk and vanishes if you do not catch it. The other is a pastoral record about a bereavement, an illness or a marriage under strain, written on a personal laptop by someone with no IT department behind them. This guide covers a practical workflow for both, with capture happening on your own device instead of a vendor's server.
Record a short reference, keep it on your Mac, and use it in supported Yaps voice workflows.
Requires macOS 13.0+ (Apple Silicon recommended)Also on iPhone and iPad