When viewers browse Prime Video, the short text summaries shown for movies and TV shows are generated by large language models rather than written by human editors. This change was introduced as part of a major redesign of the Prime Video interface in July 2024.

Amazon tapped multiple undisclosed large language models to produce synopses designed to be easier to read while browsing on a screen. This practice is distinct from X-Ray Recaps, which summarizes episodes for viewers already watching a show; these synopses are part of the content discovery experience before a viewer has chosen something to watch. The synopses may supplement or replace text descriptions previously provided by content studios.