Generate videos from text descriptions using advanced AI video models.
Overview
The Text-to-Video node creates videos directly from text prompts. Use it to generate original video content without needing source images.
Inputs
Text Prompt (required)
- Describe the video you want
- Can connect from Prompt Node or type directly
Outputs
Generated Video
- MP4 format
- 2-10 seconds duration
- Saved to Media Library
Configuration
Model Selection
Kling - High quality cinematic
Pika - Fast, good quality
AnimateDiff - Animation style
Veo - Premium quality (Pro/Enterprise)
SeeDance - Basic quality (Free plan)
Duration
- 2 seconds - Quick clip
- 5 seconds - Standard
- 10 seconds - Longer scene
Aspect Ratio
- 16:9 - Landscape
- 9:16 - Portrait
- 1:1 - Square
- Custom
Quality Settings
- Draft
- Standard
- High
- Ultra (model-dependent)
Audio Generation
Toggle on/off (some models support audio generation)
Use Cases
Action Scenes
"Superhero flying through city, dramatic chase scene, cinematic"
Nature
"Ocean waves crashing on beach at sunset, slow motion"
Abstract
"Colorful particles flowing through space, ethereal"
Product Demos
"Smartphone rotating, sleek design, studio lighting"
Prompt Tips
Structure:
[Action], [setting], [style], [mood], [camera]
Include:
- Movement/action verbs
- Camera angles/movements
- Lighting conditions
- Style references
- Time of day
Examples:
- "Dragon flying over mountains, epic, golden hour, aerial view"
- "Coffee pouring into cup, close-up, slow motion, warm lighting"
- "Neon lights reflecting on wet street, cyberpunk, night, tracking shot"
Workflow Integration
Output to:
- Video Stitcher (combine clips)
- Voice-Over Node (add narration)
- Video-to-Video (transform style)
- Editor (timeline editing)
Input from:
- Prompt Node (text)
- Chat Node (refined prompts)
Tips
Better Results:
- Be specific about camera movement
- Mention lighting and mood
- Keep prompts focused
- Generate multiple variations
Performance:
- Start with 5 seconds
- Use fast models for testing
- Final output with quality models
Cost:
- Each generation = 1 video unit
- Longer duration may cost more with overage
Troubleshooting
Poor quality:
- Use higher quality model
- Add more detail to prompt
- Try different duration
Unnatural motion:
- Simplify prompt
- Try different model
- Use image-to-video instead
Generation failed:
- Check prompt content policy
- Try different model
- Simplify prompt
Related Nodes