If you also want to support dictation, Cortana, or other speech recognition services (such as a predefined grammar defined in a topic constraint), you must also confirm that Online speech recognition (Settings -> Privacy -> Speech) is enabled. However, as the user can choose to turn this setting off at any time, you should confirm that your app has access to the microphone before attempting to use it. If the user clicks Yes to grant access to the microphone, your app is added to the list of approved applications on the Settings -> Privacy -> Microphone page. For more detail, see App capability declarations.

To automatically prompt the user with a system dialog requesting permission to access and use the microphone's audio feed (example from the Speech recognition and speech synthesis sample shown below), just set the Microphone device capability in the App package manifest. To support speech recognition with your app, the user must connect and enable a microphone on their device, and accept the Microsoft Privacy Policy granting permission for your app to use it. Speech recognition is made up of a speech runtime, recognition APIs for programming the runtime, ready-to-use grammars for dictation and web search, and a default system UI that helps users discover and use speech recognition features. Use speech recognition to provide input, specify an action or command, and accomplish tasks.

Microsoft speech to text for pc