An Insert is a close-up shot of an object, detail, or action that is edited into a sequence to provide specific visual information essential to understanding the scene or advancing the narrative. It temporarily interrupts the main action to draw attention to something that might otherwise be missed or needs emphasis for story purposes.

Inserts typically show hands performing actions, objects being manipulated, written text that needs to be read, details that reveal clues or information, or mechanical processes that require clarity. They are planned during production specifically to provide coverage that can be used during editing to clarify action, compress time, or provide visual punctuation. Classic examples include close-ups of a character's hand turning a doorknob, a clock showing the time, a letter being written, or a weapon being loaded.

In AI video generation workflows, understanding inserts helps creators plan comprehensive coverage of scenes by generating not only the main action but also detail shots that can be used for editing flexibility. While AI models may not automatically generate insert shots, prompting for specific close-up details as separate generations provides the editorial material needed for polished, complete sequences.