An open-source, fully offline voice assistant for many languages, designed for private home automation.
Rhasspy is an open-source voice assistant platform that operates entirely offline, ensuring user privacy by processing voice data locally. It allows users to define custom voice commands using a grammar-based template language and integrates with home automation systems like Home Assistant and Node-RED. The platform supports multiple languages and uses a modular service architecture for flexibility.
Savvy amateurs or advanced users who want a private voice interface for their home automation software, particularly those who are comfortable customizing software and prefer offline, open-source solutions.
Developers choose Rhasspy for its unique combination of complete offline operation, open-source licensing, and seamless integration with popular home automation platforms, offering a privacy-focused alternative to cloud-based assistants.
Offline private voice assistant for many human languages
Open-Awesome is built by the community, for the community. Submit a project, suggest an awesome list, or help improve the catalog on GitHub.
Functions entirely disconnected from the internet, processing all voice data locally to ensure user privacy, as explicitly stated in the README's key features.
Allows users to define custom voice commands using a template language that produces structured JSON events, making integration with automation software like Node-RED straightforward, as shown in the README examples.
Seamlessly works with popular platforms such as Home Assistant, Node-RED, and OpenHAB through MQTT, HTTP, and WebSockets, with built-in support highlighted in the README.
Composed of independent services communicating via MQTT using the Hermes protocol, enabling easy extension or replacement of functionality, as detailed in the services section.
Supports over 15 languages with various backends like Kaldi, Pocketsphinx, and DeepSpeech, providing flexibility for different linguistic needs, as listed in the supported languages section.
Requires manual configuration of voice commands and integration, targeting savvy amateurs or advanced users, which implies a steep learning curve and time investment for newcomers.
Optimized for pre-specified grammar-based commands; while open transcription is possible, it may not match the accuracy or ease of cloud-based assistants for free-form speech, as noted in the README.
The project is split into over 30 repositories, as shown in the development status table, which could lead to challenges in maintenance, updates, and consistency across components.
Relies on offline speech recognition backends like Pocketsphinx, which might have lower accuracy compared to cloud-based services, especially for less common languages or noisy environments.
rhasspy is an open-source alternative to the following products:
A virtual assistant developed by Google that can engage in two-way conversations and perform tasks through voice commands or text input.
Snips.AI was a privacy-focused voice assistant platform that provided offline voice recognition and natural language understanding for embedded devices, acquired by Sonos in 2019.
A cloud-based voice service from Amazon that powers devices like Echo speakers, providing voice-controlled AI assistance and smart home integration.