If you are talking about the default voice, this is set in status - hub details. There is a drop-down with the hubitat accepted voices.
To expedite, and in case you are asking something more detailed:
The command (from Common Methods documentation) is
textToSpeech(String stringToBeSynthesized, String voice = null)
I assume that any of the drop-down voices could be used in the actual command. Sadly, the default "Speak" command does not take the voice parameter - it uses the set default. (Note the speak command comes from the capability Voice Synthesis".)
You could write an app that links to a speaker with audio notification capability, do the text-to-speech in the app, and send the tts comand output to the speaker in one of the following commands: Play Track, Play Track And Restore, or Play Track And Resume. Example:
child.playTrackAndResume(textToSpeech("Hello Everyone", "Nicole")). (Nicole is the name for Nicole, Female, Australian English from the drop-down. One of my favorite voices.)