🎬 Gen-4 Video Prompting Guide

Prompting Fundamentals:

• Start simple: Begin with foundational prompts capturing only essential motion • Iterate progressively: Add elements one at a time (subject → camera → scene → style) • Simplicity is powerful: Don't underestimate basic prompts • Focus iteration: Once basic motion works, refine with additional elements

Core Best Practices:

✅ Use positive phrasing exclusively (never negative prompts) ✅ Describe MOTION, not the static image content ✅ Use general subject terms: "the subject," "the woman," "the man" ✅ Keep prompts direct, simple, and easily understood ✅ Refer to subjects with positional language for multiple people ❌ Avoid conversational or command-based prompts ❌ Avoid overly complex, lengthy descriptions ❌ Don't describe static image elements already visible

Essential Prompt Elements:

1. Subject Motion: How characters/objects move, expressions, gestures, interactions
2. Scene Motion: Environment reactions (insinuated: "dusty desert" vs described: "dust trails behind them")
3. Camera Motion: Movement style (handheld, dolly, pan, tracking shot, locked camera)
4. Style Descriptors: Motion speed, aesthetic style (live action, cinematic, smooth animation, stop motion)

Multi-Subject Prompting:

• Use clear positional language: "The subject on the left walks forward. The subject on the right remains still." • Use simple descriptive identifiers: "The woman nods. The man waves." • This helps direct specific motion without confusing the model

Scene Motion Approaches:

• Insinuated motion: "The subject runs across the dusty desert" (natural results) • Described motion: "The subject runs across the desert. Dust trails behind them" (emphasized elements)

Successful Examples:

• "The woman inspects her reflection in the mirror. The surface bubbles with translucent bubbles. Locked camera." • "The handheld camera tracks the mouse as it scurries away." • "The pile of rocks transforms into a humanoid. The rock humanoid walks around."

Generated Prompts