I've got a TTS service running on my network which I can access by hitting http://hostname:8124/synthesize/whatevertext which has integrated reasonably well with other projects, keeps the actual overhead for the speech rendering off the raspberries, and so on. I am wondering if there is an option somewhere in the guts of piper to have it call on that instead of rendering locally, or if I would need to write my own integration for it.
#Wanting to use a custom TTS
sterile cave
silk ridge
If you added support for Wyoming to that service I expect it'd work as "piper"