Glossary

The glossary of AI video.

This glossary defines 34 terms of AI-generated video and brand series, from script to release. The definitions are short and practical. They describe how the words are used in Ryze AI Studios productions.

This glossary gathers the words you come across when preparing an AI video or a brand series: writing, generation, sound, editing, release. Each term has an anchor, so you can link directly to a definition. The full method is explained on the page AI video production.

What are the terms of writing and series?

Bible: Reference document for a series or a character: world, characters, tone, rules, what never changes. The whole team writes and generates from it.
Season arc: How a story develops across all the episodes of a season: starting point, turning points, conclusion.
Vertical series: Fiction in short episodes, filmed or generated in 9:16, designed to be watched on a phone. See the guide to vertical series.
Microdrama: Fiction series with very short episodes in vertical format, built to chain twists and make viewers want the next episode.
Brand series: Episodic fiction produced for a brand, with recurring characters. The brand lives inside the story instead of interrupting it.
Script: The text of the video: structure, dialogue or voice-over, acting notes, call to action. At the studio, the client approves it before any generation.
Hook: The first seconds of a video, designed to stop the scroll: an image, a line or a question that makes people stay.
Cliffhanger: An episode ending left hanging on a question or a twist, so the audience comes back for the next one.
Branch (audience vote): A possible continuation of an interactive story. The audience chooses between two options, and the winning branch becomes the next episode. Branches are prepared in advance.

What are the terms of image and generation?

Storyboard: Visual breakdown of a video, shot by shot: framing, set, characters, movement. It fixes the visual identity before the renders.
Shot: A continuous piece of video between two cuts. In AI video, each shot is generated separately and then edited together with the others.
Art direction: All the visual choices of a project: light, palette, sets, costumes, camera style. It ensures a production has a single signature.
Prompt: Written instruction given to an AI model to describe the expected image, shot or sound.
Reference image: An approved image used as the starting point for generation: a character portrait, a set, a product. It limits drift from one shot to the next.
Character consistency: The ability to keep the same face, build, voice and outfits from one shot and one episode to the next. It relies on the bible, reference images and human selection.
Text-to-image: Generating an image from a written description.
Text-to-video: Generating a video shot from a written description alone.
Image-to-video: Generating a video shot that animates a starting image. More controllable than text-to-video, because the framing and the character are already set.
Render: The output produced by the model for a shot. Several renders of the same shot are often generated to keep the best one.
Inpainting: AI retouching of a specific area of an image: you select the area and the model regenerates it, while the rest of the image stays the same.
Outpainting: AI extension of an image beyond its edges, useful for example to go from a vertical frame to a wider one.
Upscale: Increasing the resolution of an image or video, with detail reconstruction, for display on a large screen.

What are the terms of voice, sound and editing?

Lip sync: Matching a character's lip movements to the spoken text. Essential for a talking-to-camera video.
Speech synthesis: Producing a spoken voice from text with an AI model.
Voice cloning: Creating a synthetic voice that imitates the voice of a specific person. Only to be used with that person's written consent.
Sound design: Creating the sound world of a video: ambiences, sound effects, textures, on top of the music and voices.
Colour grading: Adjusting colour, light and contrast in the edit so that all shots share a consistent look.
9:16 format: Vertical format, taller than it is wide, used by TikTok, Reels, Shorts and stories. Other common formats are 16:9 (horizontal), 1:1 (square) and 21:9 (cinema).
Version: A version of the same video adapted to another format, length, language or platform.

What are the terms of release and branding?

AI brand face: An AI-generated character who represents a brand on a recurring basis, with a face, a voice and a personality set in an identity bible. See our page about the AI ambassador.
UGC: User-generated content: a spontaneous-style video, as if filmed by a user. AI UGC is performed by a generated character, not by a real customer, and must not fake a review.
A/B test: Comparing two versions of a video that differ in only one element, the hook for example, to keep the one that performs best.
AI label: An indication that the content was generated or altered by AI: platform label, a sentence in the caption or in the video. See the page AI content.
Deep fake: Under the EU AI regulation, AI-generated or manipulated image, audio or video content that resembles existing people, objects, places or events and could falsely appear authentic (Art. 3, point 60).

Sources

Method and offers of Ryze AI Studios, accessed 30 October 2026.
Regulation (EU) 2024/1689 on artificial intelligence, Article 3, point 60: ai-act-service-desk.ec.europa.eu.

Last updated: 30 October 2026