Overview
Piper is an open-source neural text-to-speech engine that turns written text into natural-sounding speech. It is built to be fast and to run fully on your own machine, so you do not need to send text to a cloud service.
Piper uses the espeak-ng project to convert words into phonemes, then a neural voice model to produce the audio. It ships ready-made voices in many languages, works well even on small hardware like a Raspberry Pi, and is used by projects such as Home Assistant and screen readers for the visually impaired.
What it does
- Local, offline speech synthesis with no cloud dependency, so your text stays on your own device
- Fast neural voices that run on modest hardware, including small boards like the Raspberry Pi
- Pre-trained voices in many languages that you can list and download on demand
- Command-line interface, HTTP web server, Python API, and a C/C++ library for different integration needs
- Optional GPU acceleration through onnxruntime-gpu for higher throughput
- Tunable output, including volume, speaking speed, audio variation, and raw espeak-ng phoneme injection
Getting started
Install Piper from PyPI, download a voice, then generate a WAV file from text. The Python package name is piper-tts.
Install Piper
Install the Piper package from PyPI with pip.
pip install piper-ttsDownload a voice
List the available voices, then download one. This example downloads an English (US) voice into the current directory.
python3 -m piper.download_voices en_US-lessac-mediumGenerate speech from the command line
Run Piper with a voice model and write the spoken text to a WAV file. Here it creates test.wav from a short sentence.
python3 -m piper -m en_US-lessac-medium -f test.wav -- 'This is a test.'Use the Python API
You can also call Piper from Python with PiperVoice.synthesize_wav to write audio to a WAV file.
import wave
from piper import PiperVoice
voice = PiperVoice.load("/path/to/en_US-lessac-medium.onnx")
with wave.open("test.wav", "wb") as wav_file:
voice.synthesize_wav("Welcome to the world of speech synthesis!", wav_file)Commands and code are distilled from the project's own documentation — always check the official repo for the latest.
When to use it
- Adding offline voice output to home automation setups such as Home Assistant
- Powering screen readers and accessibility tools that read text aloud for visually impaired users
- Generating narration or voiceovers for videos and apps without relying on a paid cloud service
- Running a local text-to-speech web server for repeated synthesis from your own software
How Piper compares
Piper alongside other open-source audio, music & voice tools AI/TLDR tracks, ranked by GitHub stars.
| Tool | Stars | What it does |
|---|---|---|
| Whisper | ★ 112k | OpenAI's speech recognition model that transcribes and translates audio across many languages. |
| VoiceStudio | ★ 96k | A local-first desktop voice studio that clones and designs voices, dubs video, and handles dictation, running 16 text-to-speech and 11 speech-recognition engines on your own hardware. |
| GPT-SoVITS | ★ 64.3k | An open-source WebUI that clones a voice from a short audio sample and turns text into speech, with zero-shot and few-shot fine-tuning. |
| Voicebox | ★ 59.1k | Local-first voice studio that clones a voice from a short sample, generates speech across seven TTS engines and 23 languages, handles system-wide dictation, and speaks for agents over MCP. |
| VibeVoice | ★ 55.1k | Microsoft's text-to-speech model for generating long, expressive multi-speaker audio like podcasts. |
| whisper.cpp | ★ 54.8k | A dependency-free C/C++ port of Whisper built on ggml, running speech recognition on CPU, Metal, CUDA, Vulkan and NPUs from phones to servers. |
| Coqui TTS | ★ 46.2k | A library of text-to-speech models including the multilingual XTTS voice-cloning model. |
| Piper | ★ 11.3k | Fast, local neural text-to-speech that runs offline on your own machine |