Viola, open source
Read it. Run it.
Make it yours.
Viola's source core is now on GitHub. A voice assistant with desktop and browser interfaces, local music, multiroom audio, and self-hosted phone calls. Yours to explore and build on.
See how it works
Follow a voice request through the code. Inspect the tools, change a behavior, or build an integration for the way you use Viola.
Choose your AI
Use a local model or bring your own AI provider key. The source core runs without a Viola account or subscription. Your chosen providers may charge for their services.
Join in
Found a bug or have an improvement? Share a reproducible issue or a focused change. The contribution guide is a good place to begin.
Your first run
Clone viola-oss and open the SOURCE folder. You'll need Python 3.11 and Node.js 22 or later. The README walks through dependencies, the interface build, speech-model setup, and launching Viola.
See the installation guide for platform requirements and optional speech features. Use the telephony guide to connect your own carrier account.
Code you can build on
Original Viola code and commissioned artwork in the repository are published under Apache License 2.0, which allows use, modification, and redistribution under its terms.
Third-party code, models, icons, and fonts keep their own licenses. The notices explain those terms. Viola trademarks and the separate hosted service are outside the source license grant.
Looking for the desktop download?
The GitHub release is a source distribution you set up yourself. For an existing packaged Windows or Mac app, head to Downloads. Those builds and hosted services have their own accompanying terms and plans.