AI video glossary: the terms behind short dramas and product ads
AI video tools borrow words from film, advertising and machine learning. This glossary explains each term in one or two sentences, with a note on how it shows up when you make a vertical drama or a product ad.
Definitions describe the idea in general. Where AlphaDrama handles the step for you, the entry says how.
Example: Write a 60-second episode about a night-shift nurse who finds a note in her own handwriting. Hook in the first two seconds, end on a cliffhanger.
Story and structure terms
These words describe how a short story is built before any video is generated.
- Vertical drama: a serialized story shot for a phone held upright (9:16). Episodes are short, often one to three minutes, and each one ends on a reason to watch the next.
- Logline: a one or two sentence summary of the story, naming the lead, the goal and the obstacle. It is the shortest input you can give a drama generator.
- Hook: the first second or two of a video, built to stop the scroll. In a drama it is usually a conflict or a question; in an ad it is a problem or a surprising result.
- Cliffhanger: an ending that leaves a question open so the viewer continues. Episode endings are where a series keeps or loses its audience.
- Beat: one unit of story action, such as a reveal, a reversal or a confession. A scene is made of a few beats.
Planning and shot terms
These terms describe how a script becomes a list of shots.
- Storyboard: a sequence of frames or descriptions showing what each shot contains before rendering. It is where you catch a weak scene before any credits are spent.
- Shot list: the ordered list of shots with length, framing and action. A shot is one continuous piece of video between two cuts.
- Framing: how much of the subject is in the picture, such as close-up, medium or wide. Close-ups carry emotion; wide shots show where the scene happens.
- B-roll: cutaway footage that supports the main action, such as hands on a keyboard or a product on a shelf.
- Canvas: a freeform board where text, image and video nodes are connected to plan or generate shots. AlphaDrama offers one for people who prefer to work visually.
Generation terms
These terms describe how a video model produces footage.
- Text-to-video: generating a clip from a written description alone.
- Image-to-video: generating a clip that starts from a given picture, so the opening frame matches your reference.
- Reference image: a picture the model is told to follow, such as a portrait of a character or a photo of a product.
- Character consistency: keeping the same face, hair and outfit across shots and episodes. It is usually achieved by casting the character once and generating every later shot from that reference.
- Prompt: the written instruction given to a model. Short, concrete prompts about action, light and framing work better than long lists of adjectives.
- Video replica or remake: building a new video that follows the pacing and shot order of a reference clip, with your own product, person or story. No footage is reused.
Sound and finishing terms
These terms describe what happens after the picture is generated.
- Voiceover: spoken narration or dialogue added to the video, either recorded or generated by a text-to-speech voice.
- Lip sync: matching mouth movement to spoken audio so a character appears to say the line.
- Burned-in captions: subtitles rendered into the video pixels, so they show on every platform without a separate subtitle file.
- 9:16: the aspect ratio of full-screen vertical video on TikTok, Instagram Reels and YouTube Shorts. Horizontal footage is cropped or letterboxed there.
- Credits: the unit used to pay for generation. In AlphaDrama you see the credit cost of a render before you start it.
How it works
Start with a story or a product
Paste a chapter, an outline or a one-line idea. For an ad, add product photos or a product link.
We cast, storyboard and voice it
Characters get fixed looks, scenes get sets, and the script is split into shots with voices and captions.
Review the price, then export
You see the cost before any video renders. Export a 9:16 file for TikTok, Reels and Shorts.
Frequently asked questions
What is a vertical drama?
A serialized story made for phones held upright (9:16), told in short episodes that each end on a reason to watch the next one.
What is the difference between text-to-video and image-to-video?
Text-to-video starts from words only. Image-to-video starts from a picture, so the first frame matches your reference, which helps keep characters and products consistent.
What does character consistency mean?
The same person looks the same in every shot and episode. You get it by creating the character once and using that look as the reference for every later shot.
Why does an AI video need a storyboard?
A storyboard lets you fix the order and content of shots before rendering. Changing a line of text costs nothing, while re-rendering a clip costs credits.
What are burned-in captions?
Captions drawn into the video image itself. They stay visible on any platform and cannot be turned off by the viewer.
Related
Make your first video today
Sign up free and get starter credits. You see the price before anything renders.
Try this example