A to Z
AI video glossary — plain language.
Definitions of every term you'll encounter when making AI video in 2026.
AI TikTok video
A 9:16 vertical TikTok generated by an AI video model. Increasingly common on the platform in 2026.
AI video generator
Software that turns a text prompt (and sometimes an image) into a generated video clip using a trained AI model.
AI video model
A trained machine learning model that generates video. Examples include Sora 2, Veo 3.1, Kling 3.0, and Hailuo.
Aspect ratio (AI video)
The width-to-height ratio of a generated video. 9:16 for vertical, 16:9 for horizontal, 1:1 for square.
Diffusion steps
The number of iterative denoising passes a diffusion model uses to produce an output. More steps = more detail, slower.
Google Veo 3.1
Google DeepMind's photoreal AI video model. Known for faces, hands, physics, and native audio.
Image-to-video
An AI video generation method that animates a starting image — often combined with a text prompt describing the motion.
Kling 3.0
Kuaishou's smooth-motion AI video model. Best for action, dance, and choreography.
Latent diffusion
The deep learning technique behind most modern AI video models: iteratively denoising a compressed (latent) representation of a video.
Lip sync (AI video)
AI-generated video where the character's mouth movements match generated audio. Native in Google Veo 3.1.
Native audio (AI video)
Audio generated by the AI video model alongside the video — not added in post.
Prompt adherence
How faithfully an AI video model follows the details of the input prompt.
Prompt engineering (for video)
The practice of writing AI video prompts that consistently produce good outputs — through structure, camera language, and model awareness.
Seed (AI generation)
A number that controls the random starting point of a generation. Same seed + same prompt = reproducible output.
Sora 2
OpenAI's flagship text-to-video AI model. Cinematic, detailed, narrative-strong. Available inside VIBE.
Text-to-video
An AI video generation method where the input is a text prompt and the output is a generated video clip.
Vertical video (9:16)
Video shot or generated in portrait aspect ratio. The standard for TikTok, Instagram Reels, and YouTube Shorts.
YouTube Shorts (AI-generated)
9:16 vertical short-form videos posted to YouTube. AI-generated Shorts are increasingly common in 2026.