Abstract: The present disclosure discloses a method and an apparatus for controlling a virtual speech assistant, a user device and a storage medium, which solves the problem associated with bad feedback effect for input of a user device in the field. The method includes: displaying a virtual speech assistant icon in a floating way on a human-machine interaction interface of a user device; receiving a speech instruction when a microphone of the user device is enabled; and performing an operation according to the speech instruction, and producing a speech output corresponding to an operation result of the operation.