SpeakThis is made by Kazobe. For help, bug reports, or feature requests, email apps@kazobe.com. Include your device model, operating system version, SpeakThis version, and a short description of what happened. Please do not send private scripts or recordings unless you choose to share them for troubleshooting.
Download a voice model
On iPhone or iPad, choose the model marked START HERE: Pocket TTS, about 500 MB. On Mac, choose Pocket TTS, about 90 MB, or Curated Pro, about 2.9 GB. Model files come directly from Hugging Face only after you choose to download one. Keep SpeakThis open and connected to the internet until the download finishes, with enough free storage for the model.
Optional Curated Pro and Voice Designer models on mobile require an iPad with at least 8 GB of memory and about 2.3 GB of storage per model.
KittenTTS and Mobile Quality are currently unavailable for new mobile speech generation. Your existing recordings remain playable and shareable. Previously downloaded model files can still be removed in Models to free storage.
Preview and render
On iPhone or iPad, choose a voice and sample in Studio, then tap Preview Voice. Once rendering completes, Play Preview replays the saved audio without rendering again while the text and voice settings are unchanged.
On Mac, open New Prompt, choose a voice and sample in the voice sampler, and select Preview changes. Once rendering completes, use Play saved preview to replay it. You can also add scripts to a render queue and save finished WAV audio.
Privacy
Generation runs on your device after the model download. Read the SpeakThis Privacy Policy for details about local files, downloads, and support requests.
Contact
Email apps@kazobe.com for help. Describe the issue and include the device and operating system version.