TL;DR: Universal Prompt Formula

Core formula: [Subject] + [Action] + [Setting] + [Camera] + [Lighting] + [Style] + [Quality] Works on: Every major AI video tool Time to learn: 10 minutes to understand, lifetime to master Success rate: 80%+ with proper formula vs 40% random prompts Reality: This framework is foundation - personalize for your style

Every AI video tool speaks slightly different languages. But they all understand good prompts.

Here's the universal formula that works everywhere.

The Core Formula

Base Structure

[Subject] + [Action/Motion] + [Setting/Environment] + [Camera Movement] + [Lighting] + [Style/Aesthetic] + [Quality Terms]

Why this works:

• Subject: What AI focuses on

• Action: How things move (AI's job)

• Setting: Environmental context

• Camera: Perspective and movement

• Lighting: Mood and atmosphere

• Style: Visual aesthetic direction

• Quality: Output fidelity signal

All AI video tools process these components, just weighted differently.

Component Breakdown

Component 1: Subject (WHAT)

Your main focus. Be specific.

Vague:

Specific:

Rules:

• Age if person

• Key physical characteristics

• Clothing if relevant

• Distinguishing features

Component 2: Action/Motion (WHAT HAPPENS)

How subject moves or behaves.

Static descriptions:

Active descriptions:

Action verbs that work:

• Walking, running, jumping, dancing

• Rotating, spinning, revealing

• Flowing, drifting, floating

• Interacting, demonstrating, using

Component 3: Setting/Environment (WHERE)

Context and surroundings.

Generic:

Specific:

Include:

• Location type

• Key environmental features

• Atmosphere elements

• Background details

Component 4: Camera Movement (HOW FILMED)

Perspective and camera behavior.

Common movements:

Technical terms that work:

• Zoom: in/out

• Pan: left/right/across

• Tilt: up/down

• Track/dolly: following/alongside

• Orbit/circle: around subject

• Crane: up and over

• Static: no movement

• POV: point of view shot

Component 5: Lighting (MOOD)

Light quality and atmosphere.

Vague:

Specific:

Lighting descriptors:

• Time: Golden hour, blue hour, midday, night

• Quality: Soft, hard, diffused, direct

• Direction: Side, front, back, overhead

• Color: Warm, cool, neutral

• Mood: Dramatic, peaceful, energetic

Component 6: Style/Aesthetic (HOW IT LOOKS)

Visual direction and artistic style.

Generic:

Specific:

Style categories:

• Cinematic: Film-like, dramatic, storytelling

• Commercial: Clean, professional, polished

• Documentary: Realistic, authentic, journalistic

• Fashion: Editorial, high-style, trendy

• Artistic: Expressive, creative, unique

• Corporate: Professional, trustworthy, formal

Component 7: Quality Terms (OUTPUT FIDELITY)

Signal high-quality output.

Always include:

Why: Tells AI to prioritize quality over speed

Complete Example Prompts

Example 1: Product Video

Example 2: Nature Scene

Example 3: Person Scene

Example 4: Abstract

Tool-Specific Adjustments

Seedance 2.0 Specifics

Seedance excels at:

• Multi-shot sequences

• Character consistency

Optimize prompts:

[Scene 1] [Subject] in [setting], [action], [camera]
[Scene 2] Same [subject] in [different setting], [action], [camera]
[Scene 3] Same [subject], [action], [different camera angle]

Maintain: Consistency in character appearance throughout
Style: [Your aesthetic]

Sora 2 Specifics

Sora excels at:

• Physics realism

• Long continuous shots

Optimize prompts:

[Highly detailed subject description] performing [physically accurate action] in [realistic environment with specific physics], [extended camera movement description], [natural lighting]

Runway Gen-3 Specifics

Runway excels at:

• Precise motion control

• Specific camera parameters

Optimize prompts:

[Subject] [action described with specific motion parameters], [camera technical specifications: focal length, angle, movement speed], [lighting setup details], [cinematic grade/LUT]

Kling 3.0 Specifics

Kling excels at:

• Photorealism

• Extended duration

Optimize prompts:

Photorealistic [subject] [action emphasizing natural realistic movement] in [authentic real-world setting], [smooth extended camera motion], [natural realistic lighting]

Advanced Formula Variations

Minimal Prompt (When Less is More)

Sometimes simplicity works:

[Subject], [Key action], [One style term]

Example:

Use when: Simple scenes, letting AI fill in details artistically

Maximum Control Prompt

When you need precision:

[Detailed subject with every specification] performing [action broken into sub-actions] in [environment described in 3 layers: foreground, midground, background], [camera with technical parameters], [lighting broken down by source and quality], [style with reference examples], [quality with specific technical terms]

Example:

Common Prompt Patterns

Pattern 1: Transform/Before-After

[Subject] transforming from [state A] to [state B], [transformation action], [setting], [camera watching transformation], [lighting highlighting change], [style]

Example:

Pattern 2: Journey/Travel

[Subject] traveling from [point A] to [point B], [journey action], through [changing settings], [camera following journey], [lighting evolution], [style]

Example:

Pattern 3: Reveal/Discovery

[Initially hidden subject] being revealed through [reveal action], [setting with obstruction], [camera movement creating reveal], [lighting emphasizing reveal moment], [style]

Example:

Quality Keyword Library

Resolution/Clarity

• 4K quality

• 8K resolution feel

• Ultra-high definition

• Crystal clear

• Razor sharp

• Pin-sharp focus

Production Value

• Professional cinematography

• Commercial production quality

• Broadcast standard

• Film-grade quality

• Studio production

• High production value

Detail Level

• Ultra-detailed

• Intricate detail

• Fine texture

• Highly detailed

• Photorealistic detail

• Meticulous rendering

Style Quality

• Cinematic quality

• Professional polish

• Refined aesthetic

• Premium finish

• Sophisticated production

• Artisanal craftsmanship

Troubleshooting Bad Results

Problem: Output Too Generic

Cause: Vague prompt

Fix: Add 3x more specific details

Problem: Wrong Motion

Cause: Action poorly described

Fix: Use specific motion verbs

Problem: Bad Lighting

Cause: No lighting specified (AI guesses)

Fix: Describe light source and quality

Before: [No lighting mentioned]

Problem: Unclear Style

Cause: No aesthetic direction

Fix: Reference specific visual style

Prompt Testing Workflow

Iteration Process

Test 1: Baseline

Test 2: Adjust Weakest Component

Test 3: Refine

Test 4-5: Polish

Frequently Asked Questions

Does this work on ALL AI video tools?

Yes. Core components universal. May need minor adjustments per tool, but foundation works everywhere.

How long should my prompt be?

Sweet spot: 30-150 words. Shorter than 20 words = too vague. Longer than 200 words = AI may miss key points.

Should I use all 7 components every time?

Minimum 4-5 components. All 7 for best control. Adapt based on needs.

What if results still aren't good?

Keep testing. Some concepts harder for AI. Try 5-10 variations before giving up on idea.

Can I skip quality terms?

You can, but including them consistently improves output fidelity.

Do I need technical camera knowledge?

No. Basic terms like "zoom in" or "camera following" work. Technical knowledge helps but not required.

Is there a magic prompt that always works?

No magic. But this formula + practice = consistently good results.

Related Reading

• Best Seedance 2.0 Prompts: 50+ Working Examples

• Seedance vs Sora vs Kling: Tool Comparison

• Seedance Beginner Guide: First Video in 10 Minutes

www.topfreeprompts.com

Access 80,000+ prompts using this universal formula across all AI video tools. Master prompt engineering with tested templates for Seedance, Sora, Runway, Kling, and every AI video generator.