Host/run own ai2 service

I’d like to run my own app inventor service like thunkable, appybuilder, block2code, etc. How can I do it? Are there any tutorials out there, because I didn’t find any?

Thanks in advance.

1 Like

Read the Readme file


Thanks, I’ll try it.

Just be aware that there are more moving parts than just what is present in the README. That will get you to the point where you could deploy a version to Google App Engine, possibly with a build server, but doesn’t include information about running your own rendezvous server, CloudDB server, etc., and some components need additional configuration to work (i.e., Firebase, YandexTranslate).


Hi @ewpatton, are there plans on opensourcing the rendezvous server too?

We’re trying to secure our implementation with https, but running into issues connecting to since it’s not open on port 443.


Rendezvous server is already open source. It can be found at appinventor/misc/rendezvous.

Thanks @Diego I pulled and now hosting it. For some reason, I’m getting stuck at this message

Have you experienced it before? If you have, what did you do to resolve it? -

Check this topic: ¿Podrían hacer su web segura? (https)

Serving only rendezvous over https is not enough to run AI on https. You need a WebRTC connection.

If you’re running the rendezvous server over HTTPS only, the two devices won’t be able to negotiate a WebRTC connection if all the references aren’t updated. Make sure that you change the rendezvous server location to HTTPS both on the website and in the companion app.