OVOS Release Channels & Installation Options

Open Voice OS (OVOS) is a modular voice assistant platform that lets you install only the components you need. Whether you're building a lightweight voice interface or a full-featured smart assistant, OVOS gives you flexibility through modular packages and optional feature sets called extras.

To manage updates and ensure system stability, OVOS uses release channels and constraints files, allowing users to pin versions based on their desired stability level.

Choosing a Release Channel

OVOS follows semantic versioning (SemVer) with a rolling release model and supports three release channels — stable, testing, and alpha — so you can pick the right balance between cutting-edge features and system reliability.

These channels are managed via the constraints files hosted in the ovos-releases repository

1. Stable Channel (Production-Ready)

  • ✅ Bug fixes only
  • 🚫 No new features or breaking changes
  • ✅ Recommended for production or everyday use
pip install ovos-core[mycroft] -c https://raw.githubusercontent.com/OpenVoiceOS/ovos-releases/refs/heads/main/constraints-stable.txt

2. Testing Channel (Feature Updates)

  • ✅ Bug fixes and new features
  • ⚠️ Not as thoroughly tested as stable
  • 🧪 Best for early adopters or development environments
pip install ovos-core[mycroft] -c https://raw.githubusercontent.com/OpenVoiceOS/ovos-releases/refs/heads/main/constraints-testing.txt

3. Alpha Channel (Bleeding Edge)

  • 🔬 Experimental features
  • ⚠️ May include breaking changes
  • 🧪 Not suitable for production use
pip install ovos-core[mycroft] --pre -c https://raw.githubusercontent.com/OpenVoiceOS/ovos-releases/refs/heads/main/constraints-alpha.txt

💡 constraints.txt files act like version "filters". They don’t install packages directly, but ensure only approved versions get installed.

OVOS From Scratch: Custom Installation

Rather than using a full distro, you can manually pick which components to install:

This is useful if you’re building something like a Hivemind node or headless device, where you might not need audio output or a GUI.

What Are OVOS Extras?

OVOS uses Python extras (e.g., [mycroft]) to let you install predefined groups of components based on your use case.

Extra Name Purpose
mycroft Core services for full voice assistant experience
lgpl Adds optional LGPL-licensed tools like Padatious
plugins Includes various plugin interfaces
skills-essential Must-have skills (like system control, clock, weather)
skills-audio Audio I/O-based skills
skills-gui GUI-dependent skills
skills-internet Skills that require an internet connection
skills-media OCP (OpenVoiceOS Common Play) media playback skills
skills-desktop Desktop environment integrations

Full Installation Example

pip install ovos-core[mycroft,lgpl,plugins,skills-essential,skills-audio,skills-gui,skills-internet,skills-media,skills-desktop]

Minimal Installation Example

pip install ovos-core[mycroft,plugins,skills-essential]

Technical Notes

  • OVOS originally began as a fork of mycroft-core. Since version 0.0.8, it has been fully modularized, with each major service in its own repository.
  • All packages follow Semantic Versioning (SemVer), so you can rely on versioning to understand stability and compatibility.
  • Constraints files are a work in progress and won’t be finalized until the first official codename release.

⚠️ Tips & Caveats

  • Using --pre installs pre-releases across all dependencies, not just OVOS-specific ones — so use with caution.
  • You can mix and match extras based on your hardware or use case, e.g., omit GUI skills on a headless server.
  • When using constraints files, make sure all packages are pinned — it avoids installing incompatible versions.
  • After installing you need to launch the individual ovos services, either manually or by creating a systemd service

See Also