Piper
Fast, lightweight neural text-to-speech that runs offline.
⬦ Free and open source
Released under the GPL-3.0 license. There is no affiliate link on this page. The button goes straight to the project.
Free and open source, released under the GPL-3.0 license. This page carries no affiliate link. The button goes to the official project.
Get Piper →Piper is the go-to when you want speech synthesis on a small or offline device. It is quick, light, and sounds far better than the robotic TTS people expect from local tools.
Note the licensing: the actively maintained Piper is now GPL-3.0 (the older MIT rhasspy/piper repo was archived in late 2025). For most self-hosters that is fine, but check it against your project before shipping.
The good
- Fast enough for edge devices and Raspberry Pi
- Fully offline, no cloud
- Good quality for its size
The catch
- Relicensed from MIT to GPL-3.0, check your use case
- No cloning, fixed voices
- Setup is more hands-on than a SaaS
Pricing
Free and open source.
Alternatives to consider
- Kokoro-82M A tiny 82M-parameter TTS model with quality above its weight.
- OpenAI Whisper The open reference model for speech-to-text.