I am presently working on a Augmented Reality project using Vuforia that uses Speech recognition to control the objects in Unity. I was just looking for a sample working code.
Unity does not have this built in yet. They have been doing research on it for a long time and this will likely be added into Unity very soon. You can get the working Speech-to-Text(free) from the Assets store here. It is open source and you can help contribute to it if you find any problems.
As a side note, almost every OS has a Speech Recognition API. You easily make a plugin by wrapping all those API into a sing class in C# then use Unity's platform preprocessor directives to determine which one to call depending on which OS your game is running on.
Android:
SpeechRecognizer class.
see this project https://github.com/gsssrao/UnityAndroidSpeechRecognition
iOS:
SFSpeechRecognizer class
MacOS:
NSSpeechRecognizer class
Windows:
SpeechRecognitionEngine class
see this project https://github.com/LightBuzz/Speech-Recognition-Unity
Example:
class CrazySpeechRecognition
{
#if UNITY_ANDROID
Use SpeechRecognizer class
#endif
#if UNITY_IOS
Use SFSpeechRecognizer class
#endif
#if UNITY_STANDALONE_OSX || UNITY_EDITOR_OSX
Use NSSpeechRecognizer class
#endif
#if UNITY_STANDALONE_WIN || UNITY_EDITOR_WIN
Use SpeechRecognitionEngine class
#endif
}
The free Speech-to-Text from Unity is sadly discontinued, as stated in the link.