Piper is a fast, local neural text-to-speech (TTS) system originally developed by the Rhasspy / Open Home Foundation community. Active development has moved from the original rhasspy/piper repository to OHF-Voice/piper1-gpl. It is specifically optimised for the Raspberry Pi 4 and Raspberry Pi 5, generating high-quality natural-sounding speech in real time on modest CPU hardware without any cloud dependency. Voice models are distributed as ONNX files paired with a JSON config, typically 20-100 MB per voice, and are hosted on Hugging Face under rhasspy/piper-voices with dozens of languages and speakers including English, German, French, Spanish, Italian, Polish, Portuguese, Dutch, and many others, with both single- and multi-speaker variants. Piper is the default TTS backend in Home Assistant’s local voice stack, integrates with Rhasspy and Mycroft-style assistants, and is widely used in DIY humanoid and educational robotics projects that need offline speech output. Combined with Whisper for STT, Piper is the de-facto open-source pairing for fully on-device voice interfaces on droids without cloud connectivity. Licensed permissively (MIT for the original repo; GPL for piper1-gpl).
Fast, local neural text-to-speech engine optimised for Raspberry Pi 4/5 and similar low-power hardware. ONNX-based, real-time, and used in Home Assistant Voice. Voices range ~20-100 MB and run on CPU only.
Piper is a fast, local neural text-to-speech (TTS) system originally developed by the Rhasspy / Open Home Foundation community. Active development has moved from the original rhasspy/piper repository to OHF-Voice/piper1-gpl. It is specifically optimised for the Raspberry Pi 4 and Raspberry Pi 5, generating high-quality natural-sounding speech in real time on modest CPU hardware without any cloud dependency. Voice models are distributed as ONNX files paired with a JSON config, typically 20-100 MB per voice, and are hosted on Hugging Face under rhasspy/piper-voices with dozens of languages and speakers including English, German, French, Spanish, Italian, Polish, Portuguese, Dutch, and many others, with both single- and multi-speaker variants. Piper is the default TTS backend in Home Assistant’s local voice stack, integrates with Rhasspy and Mycroft-style assistants, and is widely used in DIY humanoid and educational robotics projects that need offline speech output. Combined with Whisper for STT, Piper is the de-facto open-source pairing for fully on-device voice interfaces on droids without cloud connectivity. Licensed permissively (MIT for the original repo; GPL for piper1-gpl).