Piper and Avocado both make speech on your own computer, for very different people.
Piper is a building block
Piper is a compact, quick speech engine with voices in dozens of languages. You run it from the command line, as a small web server, or from Python or C++ code. It’s made for building speech into other projects, and development continues under the Open Home Foundation.
Avocado is a studio
Avocado puts a Mac app around many voice models: pick a model and voice, type a script, generate takes, direct the delivery, clone or design voices, train one on your recordings, and organise it all in projects. No terminal needed.
The short version
Building speech into something? Piper. Making voice audio on a Mac, with lots of models to choose from? Avocado.