Today Nothing has unveiled Essential Voice, a speech-to-text engine that promises to deliver “clear, ready-to-send text in real time”. Thus, it improves upon the traditional dictation that, while turning speech into text, includes all the ums, uhs, and ahs, all of your stutters, and sometimes zero punctuation (or, perhaps even worse, wrong punctuation).
Essential Voice promises to give you “the speed and ease of talking with the clarity and flexibility of text”. It cuts filler words “for a more streamlined and considered output”. It auto-detects more than 100 languages, and lets you choose regional variants for languages like English and Spanish. Essential Voice can also translate what you’re saying while it transcribes it.
It supports shortcuts for things you say often or for specific spellings. So you can tell it to write “Nothing OS” every time it hears “nothing os”, while, for example, it can also link an address to the name of your favorite restaurant. Or it can send a transcript to your email if you just tell it to (having set this up first).
Nothing promises that Essential Voice doesn’t listen to you in the background, and only activates when you choose to use it. Once activated, your audio recording is encrypted and processed on Nothing’s servers. The generated text is then returned to your device without being stored on Nothing’s servers.
Essential Voice is now available on the Nothing Phone (3). It will make it to the Nothing Phone (4a) Pro later this month, and the Nothing Phone (4a) in early May. It integrates into the built-in keyboard and the Essential Key.
