Abstract: Gallery views in online meeting user interfaces are described. Participants may be represented on a gallery that includes a subset of participants in an online meeting. Each item in the gallery may include a video feed from a participant or a representation of a participant (a picture, an icon, etc.). A dominant speaker may be identified based on a number of factors and a history of participants' activities. Based on participation history and other factors, people may be removed or added from/to the gallery. Timestamps may be used to keep track of which participant speaks when and for how long. If no one is currently speaking, gallery spots may be filled with active participants, participants with video feeds, or at random. Participants may be enabled to select between the representation gallery, content gallery, or content views (e.g., shared data, applications, etc.).