Video Creation Guide

Master AI video generation with precision prompting and continuity techniques

Technical Specifications

Base Duration
8 seconds per generation
Maximum Length
~148 seconds (20 extensions)
Resolution Options
720p (default) | 1080p (16:9 only)
Aspect Ratios
16:9 (landscape) | 9:16 (portrait)
Audio Support
Native dialogue, ambient, SFX

Generation Modes

Text-to-Video
Create from descriptive prompts
Image-to-Video
Animate from reference images
Video Extension
Seamlessly extend existing clips
Multi-Scene Scripts
One prompt for continuous narrative
1
Write Prompt
2
Add Reference
3
Generate Base
4
Extend if Needed

Prompting for Maximum Accuracy

Example for 8-second clip: "A cinematic close-up of a woman in her 30s with dark hair, wearing a white lab coat, standing in a modern laboratory with soft blue lighting. She looks directly at the camera with a confident smile. The camera slowly pushes in. Ambient sounds of quiet lab equipment humming in the background." Negative prompt: cartoon, low quality, blurry, text overlay, dark shadows

Using Reference Images

Image-to-Video Example: Reference Image: [Portrait of a chef in a kitchen] Prompt: "The chef looks down at the cutting board and begins chopping vegetables with precise movements. Steam rises from a pot in the background. Natural kitchen sounds and knife on cutting board."

Creating 30-Second Videos with Continuity

For videos longer than 8 seconds, use the extension method to maintain visual and narrative continuity:

Step 1 - Base Clip (0-8 seconds): "A wide shot of a modern office at sunrise, golden light streaming through floor-to-ceiling windows. A young professional woman enters through glass doors, carrying a laptop bag. She walks confidently toward the camera. Ambient office sounds, footsteps on marble floor." Step 2 - Extension 1 (8-16 seconds): "The camera follows her as she walks past empty desks toward a conference room. She glances at her watch, then pushes open the conference room door. The lighting shifts from warm sunrise to cooler interior lighting. Continued footsteps, door opening sound." Step 3 - Extension 2 (16-24 seconds): "Inside the conference room, she sets down her laptop bag on the table and opens it. She pulls out her laptop and places it on the table. Camera slowly circles around her. Laptop opening sound, bag zipper." Step 4 - Extension 3 (24-32 seconds): "Close-up of her face as she looks at the laptop screen, illuminated by the display. She smiles slightly and begins typing. Camera slowly pushes in on her focused expression. Keyboard typing sounds, ambient room tone."
Pro Tips for Continuity
Match Visual Elements: Each extension prompt should reference the ending state of the previous clip (lighting, position, camera angle).

Audio Consistency: Maintain ambient sound continuity across extensions (if scene has rain, mention it in each prompt).

Camera Movement: Plan camera motion across clips (start with wide, move to medium, end with close-up).

One Script = One Call: For multi-scene narratives, write all scenes in ONE prompt with "Scene 1:", "Scene 2:", etc. The system automatically creates one continuous video.
Common Mistakes to Avoid
❌ Vague prompts like "a person walking"
❌ Calling text-to-speech separately for video dialogue (audio is automatic)
❌ Splitting multi-scene scripts into separate generation calls
❌ Using low-resolution reference images
❌ Forgetting to specify audio elements
❌ Not maintaining visual consistency between extensions