AI Video Generation Glossary: Core Terms for Beginners

Text-to-video is the core idea behind these tools: you write a description in plain words and the model generates a short video that matches it. There is no camera and no footage to edit — what you type is the input, and moving images are the output. Every other term on this list is something you control through that text.

A prompt is the text you give the model to describe what you want to see. The more concrete and visual it is — who is in the shot, where they are, the lighting, the action — the closer the result lands to what you pictured. Vague prompts produce generic video; specific, sensory ones produce specific video.

A scene is one continuous moment in a single place: one setting, one stretch of time, one main action. AI video works best when each generation stays inside one scene rather than trying to cram several moments together. In this app, the browseable catalog items are called scenes — ready-made starting points you can pick and adapt.

A frame is a single still image, and video is just many frames shown in fast sequence. When a model generates video, it produces a run of frames designed to stay visually consistent from one to the next so the motion looks smooth. The frame is also the unit a 'shot' or 'still' refers to when people talk about a particular moment in a clip.

Aspect ratio is the shape of the video — the relationship between its width and height. Tall vertical (9:16) suits phones and social feeds, wide landscape (16:9) suits screens and players, and square (1:1) sits in between. Choosing the right aspect ratio up front means the result fits where you plan to post it without awkward cropping.

Render is the step where the model turns your prompt into the actual video file you can watch and download. 'Rendering' is the brief wait while that happens, and the 'render' is the finished clip itself. If something looks off, you adjust the prompt and render again — iterating on the text is the whole creative loop.