A video calling method and video calling apparatus are provided. The video calling method includes obtaining a first video image acquired by a first terminal; performing action recognition on the first video image; and sending, in response to determining an action recognition result matches a first preset action, a first preset animation corresponding to the first preset action and the first video image to a second terminal performing video calling with the first terminal for displaying by the second terminal. With the video calling apparatus, an animation related to a scenario may be generated according to the scenario, for example, a body action of a user, provided by a video, and the animation is sent to a peer device for displaying.