Audio sensor for words on Arduino then output to ascii/text - serial/tcp

If you are using a very-limited vocabulary you can try the [u]Easy VR Shield[/u]. (I have no idea how well it works, but I'd guess you have to speak clearly and directly into the microphone and I wouldn't expect "perfection"... No voice recognition is perfect.)

with the help of bitvoicer?

And with the "help" of a PC, right?

Dragon Naturally Speaking runs on a PC or Mac. Siri and Alexa run on powerful servers.