Skip to content

Repository files navigation

Open Voice OS Installer

Your own voice assistant, on your own hardware. Open source, privacy-focused, and yours to change — on a Raspberry Pi, a Linux box, or a Mac.

Installs reported Distributions Countries

sudo sh -c "$(curl -fsSL https://raw.githubusercontent.com/OpenVoiceOS/ovos-installer/main/installer.sh)"

One command. It asks a handful of questions, then sets everything up itself — speech, skills, services, the lot.

What you can do with it

  • Ask it things. The time, the temperature, who Ada Lovelace was — and whatever else you add, because skills are open source and there are a lot of them.
  • Run your home. The Home Assistant integration turns "turn off the kitchen lights" into the thing actually happening.
  • Give it a brain. An optional LLM fallback answers what the skills do not, pointed at whichever OpenAI-compatible endpoint you like — including one you host yourself.
  • Put it in every room. HiveMind satellites share a single assistant across several devices, so the Pi in the hallway and the one in the kitchen are the same assistant.
  • Keep it yours. The configuration is a plain file on your disk, every part is open source, and the installer asks before it shares anything.

Before you start

You need curl, git, sudo, and Bash 4 or later. Macs need a little setup first.

Prefer to read the script before running it? Download it, look it over, then run it:

curl -fsSL https://raw.githubusercontent.com/OpenVoiceOS/ovos-installer/main/installer.sh -o installer.sh
less installer.sh
sudo sh installer.sh

Answer the questions with the arrow keys and Enter. Back returns to the previous screen at any point, and nothing is installed until you have answered every screen — so you can change your mind, or leave, without touching the machine.

Talk to it

The installer tells you when it is done. Then:

"Hey Mycroft, what time is it?"

Others to try: what is the temperature?, who made you?, who is Ada Lovelace?

If nothing happens, start with Troubleshooting.

No microphone? Type instead

ovos-tui-client is a terminal you can talk to OVOS through — type what you would have said, read the reply, and watch which skill answered and why. Handy before the microphone is set up, on a machine that has none, or when you want to see what the assistant is actually doing.

Talking to OVOS from a terminal

Virtualenv installs get it in the box — ~/.venvs/ovos/bin/ovos-tui and you are talking to it. Container installs run it as a container too; Running it has both.

Everyday tasks

I want to… Do this
Update Re-run the installer and answer No to "uninstall?"
Start, stop or check the services See Managing OVOS
Change settings Edit ~/.config/mycroft/mycroft.conf
Fix a microphone that mishears Calibrate it
Install without answering questions See Automation
Uninstall Re-run the installer and answer Yes to "uninstall?"

Back up ~/.config/mycroft/mycroft.conf (or ~/ovos/config/mycroft.conf for container installs) before updating or uninstalling if you want to keep your settings.

Who is running it

From installs that accepted installer telemetry, which is one of the two questions the installer asks about sharing data — this one, sent to telemetry.smartgic.io while the install runs, and separately usage metrics, sent elsewhere during normal use. Both are off unless you accept them, and Telemetry lists exactly what each one sends.

The numbers update on their own. The live dashboard has the interactive version.

Operating systems reported by installs Share of installs enabling each feature

Will it run on my machine?

Most likely yes. The installer is tested on Debian, Ubuntu and their derivatives (including Raspberry Pi OS and Linux Mint), Fedora and the Enterprise Linux family, Arch and its derivatives, openSUSE, WSL2, and macOS on both Intel and Apple Silicon.

The full list of tested versions has the details, including which combinations macOS supports.

Documentation

Screenshots

The screens the installer walks through

Rendered from the installer itself by scripts/render_screenshots.py, so they stay in step with it.

Welcome

Detected hardware

Installation method

Release channel

Profile

Features

Summary

Finish

Related projects

  • ovos-core — the assistant this installs
  • HiveMind-core — distributed voice across several devices, which this can also set up
  • raspOVOS — a prebuilt Raspberry Pi image, if you would rather flash a card than run an installer

About

Open Voice OS and/or HiveMind installer using Ansible with an intuitive and easy Text-based User Interface

Topics

Resources

Code of conduct

Contributing

Stars

319 stars

Watchers

8 watching

Forks

Releases

Sponsor this project

Used by

Contributors

Languages