JSON Prompting Guide
Advanced structured prompting for precise video control
JSON prompting lets you define your entire video generation in one structured format - prompt, mode, quality, duration, and more. It's the fastest way to repeat a setup, share configurations with your team, or create complex multi-scene videos with precise control.
Why Use JSON Prompting?
Complete Configuration
Define everything in one place: prompt, mode, quality, duration, aspect ratio, and more.
Repeatable Results
Save and reuse exact configurations for consistent video generation workflows.
Team Collaboration
Share precise prompts and settings with team members for aligned content creation.
Pro Tip: ChatGPT Integration
Ask ChatGPT for a "KreatorFlow video JSON prompt" for your idea, then copy & paste the result straight into KreatorFlow. This gives you the fastest and most accurate JSON prompts.
Example ChatGPT Request:
"Create a KreatorFlow video JSON prompt for a product commercial showing a skincare dropper with glowing effects and multiple camera angles"
JSON Prompting Demo - SunKiss Drops Commercial
Open in DriveLoading video...
Watch this 4-scene commercial created entirely with JSON prompting for precise control over lighting, camera work, and product presentation.
JSON Structure Overview
Complete Field Reference
Basic Configuration
input_type: "text" | "image" | "character"
prompt: Your video description (supports [cut])
mode: "Lite" | "Pro"
Video Settings
quality: "480p" | "720p" | "1080p"
aspect_ratio: "Wide (16:9)" | "Square (1:1)" | "Portrait (9:16)" | "Adaptive"
duration_seconds: 5 | 10
fixed_camera: true (locked) | false (dynamic)
Optional Fields
image_urlRequired when input_type = "image"
character_idRequired when input_type = "character"
Advanced Multi-Scene JSON Example
SunKiss Drops Commercial (4 Scenes)
This advanced example shows how to create a professional product commercial using structured JSON with precise scene control.
[
{
"scene": "A close-up macro shot of a dropper filled with glowing orange SunKiss Drops being pulled slowly from the bottle.",
"action": "The liquid glistens and clings to the glass dropper as it rises. Light flares subtly in the background.",
"camera": "Extreme close-up, shallow depth of field, glowing backlight.",
"duration": "2.5 seconds"
},
{
"scene": "The dropper dispenses a single drop of the glowing liquid into a palm with glowing sun-kissed skin.",
"action": "The drop lands in slow motion and gently spreads as fingers prepare to rub it in.",
"camera": "Close-up, side angle with natural warm lighting and soft skin focus.",
"duration": "2.5 seconds"
},
{
"scene": "A glowing model in their 20s with radiant tan skin applies SunKiss Drops to their cheekbone, smiling slightly.",
"action": "They gently tap the drops in with their fingertips, showing a subtle glow appearing on the skin.",
"camera": "Medium beauty shot with golden hour light washing over the face.",
"duration": "2.5 seconds"
},
{
"scene": "The product bottle rotates slowly on a glossy surface with sunlight streaks and soft sparkles around it.",
"action": "The text on the bottle glows briefly, emphasizing 'SunKiss DROPS'.",
"camera": "360-degree product shot with reflection on the surface, soft cinematic focus.",
"duration": "2.5 seconds"
}
]Advanced Techniques Used:
- • Precise timing: Each scene exactly 2.5 seconds for perfect 10s total
- • Lighting consistency: Glowing/warm lighting theme throughout
- • Camera progression: Macro → Close-up → Medium → Product shot
- • Action continuity: Logical flow from product to application to result
- • Brand focus: Multiple product reveals and brand emphasis
Simple JSON Examples
Basic Text-to-Video JSON
{
"input_type": "text",
"prompt": "Cinematic shot of woman walking through neon-lit city at night",
"mode": "Pro",
"quality": "1080p",
"aspect_ratio": "Wide (16:9)",
"duration_seconds": 5,
"fixed_camera": false
}Result:
5s Pro 1080p wide video with dynamic camera = 25 credits
Character Animation JSON
{
"input_type": "character",
"character_id": "char_123",
"prompt": "My model waves and smiles at camera",
"mode": "Lite",
"quality": "720p",
"aspect_ratio": "Portrait (9:16)",
"duration_seconds": 5,
"fixed_camera": true
}Result:
5s Lite character video with fixed camera = 7 credits
JSON Credit Quick Reference
Credit Calculation Formula
Text/Image → Video:
Lite 480p 5s=5, 10s=10
Lite 720p 5s=10, 10s=20
Lite 1080p 5s=20, 10s=40
Pro 480p 5s=10 / 10s=15
Pro 720p 5s=15 / 10s=25
Pro 1080p 5s=25 / 10s=35
Character → Video:
Add +2 credits to above values
Example: Lite 720p 5s = 10+2 = 12 credits
Example: Pro 720p 5s = 15+2 = 17 credits
💡 Quick tip: Include these values in your JSON planning to stay within budget limits.
ChatGPT Integration Examples
Example Requests for ChatGPT
Product Commercial:
"Give me a KreatorFlow video JSON prompt for a product on a table with commercial studio lighting."
Results in professionally structured JSON with proper lighting and camera angles.
Multi-Scene Character:
"Create a multi-scene JSON for Image → Video where a model turns → waves → walks forward."
Generates structured scenes with smooth action transitions.
Cinematic Sequence:
"KreatorFlow JSON prompt for a cinematic car chase scene with multiple camera angles."
Creates dynamic multi-angle sequences with proper cinematic flow.
JSON Prompting Best Practices
Optimization Tips
Start Simple:
Get one clear action working before adding complexity.
Use [cut] in JSON:
Combine JSON structure with [cut] syntax for multi-scene control.
Be Action-Driven:
Focus on "turns," "waves," "walks," "camera pushes in" for clarity.
Advanced Techniques
Pick a Style and Stick:
Maintain consistent visual style per video for smoother transitions.
Test with Lite First:
Validate complex JSON prompts with Lite 5s before Pro 1080p.
Save Templates:
Create reusable JSON templates for common video types.
Master Structured Video Creation
Now that you understand JSON prompting, combine it with other advanced features:
