AI Video Glossary

From diffusion models to prompt engineering -- every term you need to know about AI video creation.

9

9:16 Aspect Ratio

The vertical video format (1080x1920 pixels) used as the standard for TikTok, Instagram Reels, and YouTube Shorts, filling the entire mobile screen.

A

AI Avatar

A digitally generated human character used as a consistent on-screen presenter in AI-generated videos, replacing the need for a real person on camera.

AI Captions

Automatically generated subtitles and on-screen text for video content, created using speech recognition AI like Whisper and rendered in styled formats like ASS.

AI Content Disclosure

The practice and legal requirement of labeling video content as AI-generated, following platform policies from TikTok, YouTube, Instagram, and emerging regulations.

AI Trend Discovery

Using AI tools to identify trending topics, viral content patterns, and content gaps within a niche to inform video creation strategy.

AI Video Script

A script generated by AI for short-form video content, typically following a hook-story-CTA structure optimized for platforms like TikTok and Reels.

C

CogVideoX

An open-source text-to-video and image-to-video diffusion model by Tsinghua University, capable of running locally on consumer GPUs with 6-12 GB VRAM.

E

Edge TTS

Microsoft's free text-to-speech service offering high-quality neural voices in 400+ voice options across 100+ languages, used for AI video voiceovers.

F

Faceless Channel

A social media video channel that produces content without showing the creator's face, using AI avatars, B-roll footage, text overlays, or AI-generated visuals instead.

I

Image-to-Video (I2V)

AI technology that animates a still image into a video clip, preserving the original visual style while adding realistic motion and camera movement.

L

Lip Sync

AI technology that automatically matches a video character's mouth movements to audio speech, creating the appearance of natural talking.

P

Prompt Engineering for Video

The practice of crafting precise text descriptions to guide AI video generation models toward producing specific visual outcomes, camera movements, and styles.

R

Runway Gen-4

Runway's fourth-generation AI video model designed for professional creative workflows, offering strong camera control, style consistency, and fast generation.

S

Short-Form Video

Video content under 60 seconds, typically in vertical 9:16 format, designed for platforms like TikTok, Instagram Reels, and YouTube Shorts.

Sora 2

OpenAI's advanced video generation model capable of producing up to 20 seconds of high-fidelity 1080p video from text prompts or reference images.

T

Text-to-Video (T2V)

AI technology that generates video clips directly from written text descriptions, turning prompts into moving visuals without cameras or footage.

Token-Based Pricing

A pay-per-use pricing model for AI platforms where users purchase tokens that are consumed when generating videos, scripts, or other AI content.

V

Veo 3

Google DeepMind's video generation model that produces high-quality video with synchronized audio, available through Kie.ai and Google's Vertex AI.

Video Diffusion Model

A type of generative AI model that creates video by iteratively removing noise from random data, guided by text or image prompts to produce coherent motion.

Video Generation Pipeline

The end-to-end automated workflow for AI video production: from script generation through voice synthesis, video creation, caption overlay, to publishing.