JSON Prompt Gen Documentation

Everything you need to know to create cinematic video prompts with professional precision

Quick Start Guide

Your First Cinematic Prompt in 60 Seconds

1 Create Your Scene

Click the "+" button in the Prompt Generator tab to add your first scene. This opens the scene editor where you'll define your visual narrative.

2 Craft Your Description

In the main text area, describe your scene with cinematic specificity. Instead of "a car driving," try "A vintage red convertible speeding along a coastal highway at sunset, ocean waves crashing below."

3 Enhance with Professional Parameters

Use the parameter pills to refine your vision:

  • Camera: Select movements like "Dolly Zoom," "Eye Level," or "Aerial View"
  • Lighting: Choose moods such as "Golden Hour," "Neon Cyberpunk," or "Backlit"
  • Negative: Add or select negative keywords like "Unrealistic," "Blurry," or "Deformed"
  • Parameter: Set project-wide settings (aspect ratio, resolution, frame rate, target model) that apply to all scenes unless individually overridden

4 Generate Your JSON

Click the "Generate JSON" button in the Prompt Generator tab. The system instantly compiles your scene into a structured JSON format optimized for the choosen AI video platform.

5 Export and Execute

Use the clipboard icon to copy your JSON directly, or you can simply download it as a file. Paste this into your preferred AI video generation tool (Veo, Sora, Runway, Luma, Kling) to bring your vision to life.

Core Concepts

Multi-Scene Narrative Construction

JSON Prompt Gen transforms complex storytelling into manageable components through its scene-based architecture. Each scene operates as an independent cinematic unit while contributing to your overall narrative flow.

Scene Anatomy

Every scene contains four essential elements that work together to create professional results:

Subject & Action

The core narrative element describing who/what is the focus and what they're doing

Camera Choreography

Precise camera movements and positions that guide the viewer's eye

Lighting Atmosphere

Mood-setting illumination that transforms visual storytelling

Negative Guidance

Elements to exclude (blurry details, unwanted artifacts, style drift)

Version Control & History

Professional filmmaking requires iteration. Our history system ensures you never lose creative progress:

  • Per-Scene Undo/Redo: Each scene maintains its own history stack, allowing granular control
  • Checkpoint Restoration: Create named checkpoints for major milestones and restore them instantly

✅ Best Practice

Create a checkpoint before experimenting with radical changes to your prompt structure.

AI Enhancement Workflow

Bring Your Own Key (BYOK) Architecture

JSON Prompt Gen respects your privacy and budget control through its zero-server architecture. Your API keys and prompts remain on your device, never transmitted to third-party servers.

Setting Up Your LLM Connection

  1. Access Integration Panel: Click the AI assistant icon (robot) in the sidebar menu or click "LLM Integration" text in the top navigation bar
  2. Select Provider: Choose from OpenAI, Anthropic (Claude), or Google Gemini based on your needs and existing subscriptions
  3. Enter Credentials: Input your API key securely—keys are encrypted with AES-256 before storage in your browser
  4. Configure Settings: Set default parameter model version for consistent results

AI Enhancement Features

Prompt Expansion

Use Enhance button to Transform simple descriptions into rich cinematic narratives with professional terminology

Quality Optimization

Use Enhance button for refinement of prompts to reduce common AI video generation issues

🔒 Security Assurance

All API communication occurs directly between your browser and the LLM provider. We implement client-side encryption and never store your keys on servers. A comprehensive audit log tracks all AI interactions for transparency.

Cinematic Dictionary

Your Cinematic Knowledge Base

The Cinematic Terms Dictionary transforms technical filmmaking terminology into accessible, bridging the gap between creative vision and technical execution.

Browsing the Dictionary

Access the dictionary through the dedicated tab in the main navigation. Content is organized into intuitive categories:

  • Camera Techniques: Shot types (Close-up, Wide shot), Movements (Pan, Tilt, Handheld), and Angles (High Angle, Dutch Angle)
  • Lighting Setups: Quality (Hard, Soft), Direction (Front, Side, Back), and Mood (Suspenseful, Hopeful, Whimsical)
  • Color Theory: Palettes (Complementary colors, Analogous colors), Grading Styles (Teal and Orange, Monochromatic), and emotional impact
  • Composition Principles: Visual rules (rule of thirds, leading lines), depth creation (foreground/middleground/background), and balance techniques
  • Mood & Atmosphere: Descriptive terms that convey emotional tone and narrative context

Interactive Learning

  • Usage Examples: See how terms integrate naturally into prompt structures
  • One-Click Copy: Instantly copy any term to your clipboard for immediate use in your prompts
  • Community Contributions: Add your own definitions, examples, or categories to expand the knowledge base

Frequently Asked Questions

Privacy & Security

Q: Is my data truly private?

A: Absolutely. JSON Prompt Gen operates entirely within your browser. No prompt content, API keys, or personal data is transmitted to external servers unless you explicitly use the feedback form. Your API keys are encrypted with AES-256 encryption before storage in your browser's local storage.

Q: How are API keys protected?

A: Keys are encrypted using device-specific fingerprints and stored only in your browser's local storage. We use industry-standard encryption practices with zero-knowledge architecture—meaning even we cannot access your keys.

Technical Compatibility

Q: Which AI video platforms support this JSON format?

A: Our output is optimized for advanced platforms including Google Veo, OpenAI Sora, Runway Gen-3, and Luma Labs. Each platform has specific schema requirements, and our generator automatically formats your prompts accordingly based on your selected target platform.

Q: Can I use this offline?

A: Yes! As a Progressive Web App (PWA), JSON Prompt Gen can be installed on your device and used completely offline for prompt creation and editing. AI enhancement features require internet connectivity to communicate with LLM providers.

Workflow Optimization

Q: How do I create consistent character appearances across multiple scenes?

A: Use the Character Bible feature (available in PRO tier) to define character details once, then reference them across scenes. For the free version, include consistent descriptive details in each scene's negative prompts to minimize drift.

Q: What's the most effective way to fix common AI video issues?

A: Leverage negative prompts strategically: for morphing issues, add "consistent character design, stable anatomy"; for blurry outputs, include "sharp focus, high detail, 8k resolution"; for unwanted elements, specify exactly what to exclude.

Getting Support

Q: How do I report bugs or request features?

A: Use the built-in feedback form accessible through the help menu. For urgent issues, check our community forum where our development team actively monitors and responds to user reports.

Q: Are there video tutorials available?

A: Yes! Visit our documentation portal for step-by-step video guides covering everything from basic prompt creation to advanced cinematic techniques. New tutorials are added monthly based on community requests.

Professional Tips for Maximum Impact

Cinematic Prompt Engineering Best Practices

The Three-Layer Approach

Structure your prompts with three distinct layers:

  1. Subject Layer: Who/what is the focus
  2. Action Layer: What is happening dynamically
  3. Atmosphere Layer: How it feels emotionally and visually

Parameter Hierarchy

Prioritize your parameters in this order for optimal results:

  1. Camera movement and position
  2. Lighting quality and direction
  3. Color palette and grading
  4. Composition rules
  5. Negative guidance

Platform-Specific Optimization

Veo

Emphasize camera movement details and temporal continuity

Sora

Focus on physics accuracy and material properties

Runway

Prioritize stylistic consistency and artistic direction

Luma

Optimize for high-fidelity motion and naturalistic lighting transitions

Kling

Focus on complex human-centric actions and temporal stability