An agent apparatus includes: a microphone which collects audio in a vehicle cabin; a speaker which outputs audio to the vehicle cabin; a recognizer which recognizes the meaning of audio collected by the microphone; a display provided in the vehicle cabin; and an agent controller which causes the display to display an agent image in a form of speaking to an occupant and causes the speaker to output audio spoken by the agent image, wherein, when an image provided to the occupant is displayed on the display, the agent controller generates an agent image for guiding the line of sight of the occupant to a display position of the provided image.