With windows, the speech to text feature is enable on the OS layer (with sound settings), then OneNote is able to use it (with specific settings too)
With android, as we can say ‘ok google’ to activate the voice recogn. I imagined there is not so many thing to do, to use the sound and produce the output of that (like @voice aloud reader does)

the only thing I found with react navtivbe is this project. Others projects use TTS (TextToSpeech) instead of STT.