OpenVoiceOS
An open-source voice assistant platform for Linux, Raspberry Pi and Docker, with offline speech plugins and support for local OpenAI-compatible LLM servers.
OpenVoiceOS is a free, open-source voice AI platform for people building their own smart speakers or adding voice control to devices. Its core builds on a fork of MycroftAI/mycroft-core, and most classic Mycroft skills also work with it. The project uses the Apache 2.0 license, which permits personal and commercial use.
It runs on Linux desktops and laptops, Raspberry Pi, Mycroft Mark I and Mark II devices, and in Docker. A dedicated Linux image supports embedded hardware, including headless devices and small touchscreens. Customizable wake words and a customizable UI let builders adapt the assistant to their hardware.
Skills handle tasks such as smart home control, music and radio playback, questions, timers and reminders. You can add community skills or write your own. For conversations beyond what those skills can handle, ovos-persona connects the assistant to an LLM, including through a local OpenAI-compatible server or an OpenAI plugin.
The core runs locally without a backend internet server, but the default speech recognition service needs an internet connection. Offline recognition options include Vosk and Mozilla DeepSpeech, while Mimic3 provides local text-to-speech. Choosing local speech plugins avoids sending that processing to online services. Skills and other plugins may still need internet access; cloud speech choices include Google TTS and Amazon Polly.
