Take a quick look at Unreal Tournament 2004. I’m sure other games must use this technique, but UT2004 is the only one I kno for certain.
UT uses windows own speech to text features and then interprets the text. There’s no need for the game programmers to try and reinvent the wheel becuase the OS takes care of the interpretation, and the game need only work with its own custom library…..
At least thats how it seems to work. Improving your windows voice profile certainly improves the responsiveness of the game.