# Subwave

Subwave is a self-hosted, AI-powered personal internet radio station that streams one shared live broadcast from your Navidrome or Subsonic library, with an AI DJ selecting tracks and speaking station idents, time. It is open source and self-hosted, installable on Unraid.

Canonical page: https://ourlocalhost.com/app/3363/subwave

## Description

Subwave is a self-hosted, AI-powered personal internet radio station that streams one shared live broadcast from your Navidrome or Subsonic library, with an AI DJ selecting tracks and speaking station idents, time checks, weather, and song intros for homelabbers who want a shared radio experience rather than personal playlists.

## Platform availability

Available on 2 of 10 platforms: Unraid, Any server.

## Desktop and other devices

Other (ios,android).

## Health score

68/100 (Good), based on 4 of 5 factors.
- Maintained: 100 (last commit 5 days ago)
- Popular: 64 (1,665 GitHub stars)
- Easy to install: 55 (3 install platforms, install notes, Docker image)
- Security check: 20 (1 serious finding, 19 minor, CI, tests)

## Security check

1 serious finding: Uses dompurify 3.4.16, which has a known critical vulnerability.
Automated checks on commit 3eddf63, 25 Oct 2026. Not an audit.

## Links

- Official website: https://www.getsubwave.com
- desktop: https://www.getsubwave.com/
- Documentation: https://www.getsubwave.com/manual
- Unraid: https://ca.unraid.net/apps/subwave-073qgwu0ch9rtu?q=subwave
- GitHub: https://github.com/perminder-klair/subwave

## Common questions

### Do I need to run a separate reverse proxy in front of the web player?

The all-in-one build includes a Caddy front-end. Icecast, Liquidsoap, the DJ controller, and the web player run in one container on one port.

### Where should I point backups for settings and recorded audio?

Your settings, library cache, and recordings live under the Appdata path.

### Does the AI DJ require a hosted language model?

It uses a language model, with a local Ollama box as the default. You can also use any hosted provider you choose.

### Do I need to install a separate TTS voice before the DJ can talk?

No, the built-in Piper voice can talk straight away with nothing else to install.

## Alternatives

- VibeSesh

---
Source: Our Local Host (https://ourlocalhost.com/app/3363/subwave), the self-hosted and open-source app directory.
