piper-tts

What it is and what it does

Piper TTS is a neural text-to-speech system designed for offline, local synthesis without cloud dependencies. It uses onnxruntime for inference and embeds espeak-ng for phonemization, supporting multiple languages and voices. The package provides a Python API, command-line interface, and HTTP server, making it suitable for integration into applications like Home Assistant, NVDA screen readers, and voice assistants.

The project is actively maintained by the Open Home Foundation and has been adopted by numerous open-source projects and accessibility tools. It trades off some quality for speed and local execution, positioning itself as a practical alternative to cloud-based TTS for developers who need on-device speech synthesis without external API calls or licensing overhead.

Use it for:

• Add speech output to Home Assistant automations and voice assistants without cloud dependencies
• Enable text-to-speech in accessibility tools like screen readers (NVDA) for visually impaired users
• Synthesize narration for video processing or content generation pipelines running locally
• Embed speech synthesis in IoT or embedded systems (Jetson, edge devices) with low latency
• Build multilingual chatbots or voice interfaces that operate entirely offline

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Piper TTS is a local neural text-to-speech engine that converts text to speech using embedded phonemization, with support for multiple languages and voices.

Yes, if you need local, offline text-to-speech without cloud APIs. The package is actively maintained, widely adopted in production (Home Assistant, NVDA), has no known vulnerabilities, and supports modern Python versions. The GPL-3.0 copyleft license is a consideration for commercial use. Medium install friction is manageable given the availability of pre-built wheels across platforms.

Install

piper-tts on PyPI

pip install piper-tts

uv add piper-tts

poetry add piper-tts

Installing piper-tts

Before you install

Medium install friction due to compiled wheels for multiple platforms (macOS x86/ARM, Linux x86/ARM, Windows). Active maintenance with recent release (1 day old) and strong community adoption (5130 stars). Requires Python 3.9+.

License in practice

GPL-3.0-or-later (copyleft): any derivative work or distribution must also be open-source under GPL-compatible terms. Suitable for open-source projects and internal use, but requires careful review before commercial deployment.

Quickstart

pip install piper-tts

from piper.voice import PiperVoice

voice = PiperVoice.load('en_US-lessac-medium')
voice.synthesize('Hello world', 'output.wav')

Requires onnxruntime and pathvalidate as runtime dependencies; espeak-ng is embedded for phonemization.

Verify before relying

• Whether pre-trained voice models are included or must be downloaded separately
• Typical latency and throughput characteristics for real-time speech synthesis
• Memory footprint and GPU acceleration support via onnxruntime

Package facts

Evidence: piper_tts-1.6.1-cp39-abi3-macosx_10_9_x86_64.whl; piper_tts-1.6.1-cp39-abi3-macosx_11_0_arm64.whl; piper_tts-1.6.1-cp39-abi3-manylinux_2_17_aarch64.manylinux2014_aarch64.manylinux_2_28_aarch64.whl; piper_tts-1.6.1-cp39-abi3-manylinux_2_17_x86_64.manylinux2014_x86_64.manylinux_2_28_x86_64.whl; piper_tts-1.6.1-cp39-abi3-win_amd64.whl

Keywords: home, assistant, tts, text-to-speech

Tags

More Speech packages