When to choose Piper
Choose Piper when a particular speaker and accent matter more than having many voices in one download.
- Model type
- One speaker per download
- Curated downloads in Outmake
- About 67–116 MB each
- Languages
- Depend on the voice
- Samples below
- Published Piper samples
- Voice licences
- Vary by voice
Where shown, download sizes are rounded from the app’s pinned packages. They exclude the shared speech engine and do not measure runtime memory.
Listen before downloading
Cori is a British English voice. Kristin is an American English voice. The samples below come from the Piper sample collection, not a new benchmark in Outmake.
Both model cards identify public-domain LibriVox recordings as their training data. Other Piper voices have different terms. Check the card for the exact voice and quality level you choose.
Synthetic speech from the Piper sample collection. Both cards identify public-domain LibriVox training data. These are published samples, not recordings made in Outmake. The text reads from Wikipedia’s Rainbow article (CC BY-SA).
Each voice is its own model
With Piper, changing speakers usually means downloading another model. Outmake’s curated list includes Jenny, Alan, Lessac, Cori, Ryan, Kristin, Northern and Amy.
The package names include a language, a speaker and a quality level. Cori high and Kristin medium are different packages. The quality word is the package’s label; it is not a head-to-head listening score.
Read the licence on the individual voice
Piper is not one blanket licence covering every voice. In Outmake’s curated list, some voices permit commercial use, some are marked non-commercial, and some have no stated licence.
For a private read-aloud task, choose a voice you find comfortable. Before publishing narration or using it commercially, follow the voice’s model card and source-data terms. An unclear licence is a reason to choose another voice.
Run Piper in Outmake
Outmake is free and requires macOS 26 or later on Apple silicon. These voice models run locally after their files have downloaded. You can also follow the upstream project links below to use another runtime.
- Open Models, then Voices.
- Open the Piper voice you want. Check the licence shown on its card.
- Choose Get. Once downloaded, play a sentence in the voice’s page.
- Choose Use to make it the local voice. To hear a reply, open its menu and choose Read Aloud.
Voice settings and read-aloud help explain how to choose voices and stop playback. This page is about generating speech. For speech-to-text, use dictation or transcription instead.
Before you rely on it
- A voice’s package size does not measure its speed or runtime memory.
- A sample demonstrates the voice, not the exact output of every Piper runtime or version.
Try a short passage before committing to a long one. Local speech keeps synthesis on your Mac; it does not make an online chat model local. Outmake’s privacy policy explains the difference.
Sources
Model cards and runtime documentation checked 11 November 2026. App-specific details were checked against Outmake’s voice catalog and runtime configuration. No new listening test or speed benchmark was performed for this guide.