Using Assist / tts within Node Red with an Atom Echo

Does anyone have any experiance sending voice commands to Assist or using tts using Node Red for an Atom Echo? I was able to catch the voice commands using the sentence node. But when I want to for example send a conformation of a command received in speech back to the Atom Echo i am a bit at a loss how. Using the tts service I cannot seem to find anything specific to send the text command too. I suspect I should not try and send it to the Atom Echo but to Assist? But how if that is the case?
I also use other smart speakers like Google Mini’s and there you can send a tts command to the speaker itself.

I don’t really have any node examples to share. All i currently have is a very basic setup with one sentence node to pickup the command and one call service node to try and send a response back.

Look in a call service, the domain drop down. Do you have tts there?

Yes… And this works with other speakers (I want to see if I can move from my google speakers to using native HA with Atom speakers and others). However sending tts to the atom nothing works…

With my google Mini’s this works.
Screenshot 2024-03-24 at 12.13.50

With the Atom, nothing happens.
Screenshot 2024-03-24 at 12.14.59

Try speak maybe. Also try from dev tools in HA. I assume the atom shows up as a media_player? Can you play music through it? In other words are you sure it is working?