AI Video Prompt Template: Script, Scenes, Voiceover, and Captions

The short answer

Use a scene-level template. One giant prompt usually produces inconsistent clips. A reliable prompt describes subject, action, camera, setting, style, duration, and what must not appear.

The reusable template

For each scene, fill: goal, subject, action, camera movement, setting, lighting, duration, caption, negative constraints, and edit note. Keep style consistent across scenes.

Example scene prompt

“Scene 1: A clean desk with a laptop and printed checklist. Slow push-in camera. Bright natural office lighting. 4 seconds. Caption: ‘Start with the brief.’ No logos, no readable brand names, no real people.”

Output quality checklist

Look for visual consistency, readable captions, correct aspect ratio, no unwanted logos, no misleading realism, and a clear next step.

How to make variants

Change only one variable at a time: hook, platform, audience, or visual style. If you change everything, you will not know what improved the result.

Scene prompt grid visual for subject action camera style duration notes
Original WebP checklist visual for this workflow.
Prompt fieldWhat to writeWhy it matters
SubjectObject or scene focusPrevents random visuals
ActionWhat happens on screenCreates motion
CameraStatic, push-in, top-downControls feel
Negative constraintsNo logos, no real peopleReduces risk

Save this workflow: Before publishing, copy the checklist into your own brief, add real screenshots, and link the article to the most relevant Office, Video, or SEO hub.

FAQ

Should prompts mention a tool name?

Only if the tutorial is about that tool. Generic generated visuals should stay unbranded.

How many scenes for a short video?

Three to six scenes is enough for a 20-45 second tutorial clip.

Can I reuse this for ads?

Yes, but add stricter claims review and platform policy checks before publishing.