Google Veo 3 Prompting Guide

For high-quality video generation in Google Veo 3, prompts should follow a specific 7-component architecture to ensure technical precision and creative excellence. This structure covers visual, auditory, and technical requirements.

Prompt Components:

1. Subject: Detailed character/object description (aim for 15+ specific physical attributes, clothing, age, build, facial features, ethnicity, hair, eyes, posture, mannerisms, emotional state).
2. Action: Specific movements, gestures, behaviors, timing, sequences, transitions, micro-expressions, and body language.
3. Scene: Detailed environment (location, props, background, lighting setup, weather, time of day, architectural details, furniture).
4. Style: Camera specifications (shot type, angle, movement, lighting style, visual aesthetic, aspect ratio, film grade, color palette, depth of field).
5. Dialogue: Character speech with emotional tone indicators, pacing, volume, and accent.
   • Format: (Character Name): "Exact dialogue here" followed by (Tone: emotional descriptor/delivery style).
6. Sounds: Specific audio elements (ambient sounds, effects, background audio, music, environmental noise, natural acoustics).
7. Technical (Negative Prompt): Elements to avoid (subtitles, captions, watermarks, text overlays, unwanted objects, poor quality, artifacts).

Dialogue Optimization Tip:

For the 8-second Veo 3 format, optimize dialogue to be approximately 12-15 words or 20-25 syllables.

Subject: [Detailed character/object description with 15+ specific physical attributes, clothing, age, build, facial features, ethnicity, hair, eyes, posture, mannerisms, emotional state]

Action: [Specific actions, movements, gestures, behaviors, timing, sequence, transitions, micro-expressions, body language, interaction patterns]

Scene: [Detailed environment description including location, props, background elements, lighting setup, weather, time of day, architectural details, furniture, decorations]

Style: [Camera shot type, angle, movement, lighting style, visual aesthetic, aspect ratio, film grade, color palette, depth of field, focus techniques]

Dialogue: [Character speech with emotional tone indicators, pacing, volume, accent, speech patterns]
(Character Name): "Exact dialogue here"
(Tone: emotional descriptor/delivery style)

Sounds: [Specific audio elements including ambient sounds, effects, background audio, music, environmental noise, equipment sounds, natural acoustics]

Technical (Negative Prompt): [Elements to avoid - subtitles, captions, watermarks, text overlays, unwanted objects, poor quality, artifacts]

To ensure Veo 3 correctly interprets camera positioning, you must include the explicit phrase "(thats where the camera is)" when specifying where the camera is located. Generic viewpoint terms are often insufficient; this phrase triggers camera-aware processing.

Expert Templates

• Professional/POV: "Expert [PROFESSION] is holding a phone at arm's length (thats where the camera is) in [RELEVANT WORKSPACE]..."
• Eye Level: "POV shot from the camera positioned at eye level (thats where the camera is) as the character explains the process."
• Over-the-shoulder: "Over-the-shoulder view with camera behind the interviewer (thats where the camera is), focusing on the subject's responses."

Comparison

• ❌ Poor: "Close-up shot of the chef cooking"
• ✅ Expert: "Close-up shot with camera positioned at counter level (thats where the camera is) as the chef demonstrates knife techniques"

Handheld camera held at chest height (thats where the camera is) tracking the subject as they walk through the market

Maximize video quality by using a cross-modal reinforcement framework. Instead of treating text, audio, and visuals as separate entities, synchronize them to create a cohesive experience.

Framework Layers:

• Visual Layer: Defines the subject, lighting, camera movement, and specific visual details (e.g., microscopic views, diagrams).
• Audio Layer: Defines narration, ambient sounds, and specific 'activity' sounds that match visual cues.
• Text Layer: Defines visual overlays or key terms that appear in sync with the narration.

Cross-Modal Connections:

• Synchronize narration with visual zooms (e.g., when a narrator says a specific term, the camera zooms into the corresponding object).
• Match audio intensity to visual changes (e.g., forest sounds intensify when discussing the environment).
• Animate text overlays in sync with the process being explained.

Integrated Prompt: Macro shot of a green leaf with sunlight filtering through, camera slowly zooms into cellular level revealing chloroplasts. Animated molecular diagram shows CO2 + H2O + light energy → glucose + oxygen. Professional educational lighting, smooth zoom transitions. Audio: Clear female narrator explains: 'Photosynthesis converts sunlight into chemical energy.' Subtle forest ambiance, gentle molecular 'transformation' sounds. Key terms 'Chloroplast,' 'Glucose,' 'Oxygen' appear as clean text overlays synchronized with narration. No distracting background music.

The guide provides four advanced reasoning frameworks for complex video planning:

1. Chain-of-Thought (CoT): A step-by-step reasoning process where you define the learning objective, narrative structure, visual elements, audio strategy, and technical specs before generating the final prompt.
2. Tree-of-Thought (ToT): Exploring multiple creative branches (e.g., Traditional vs. Lifestyle vs. Problem-Solution) to evaluate pros/cons before selecting the optimal path for the final prompt.
3. Graph-of-Thought (GoT): Mapping interconnected storylines using a vertex network (characters/elements) and graph operations (Merge, Transform, Validate, Optimize) to create complex, multi-character scenes.
4. Self-Consistency: Generating multiple different approaches to the same concept (e.g., Transformation vs. Community vs. Personal Challenge) and selecting the most consistent emotional arc and message for the final optimized prompt.

To create complex, storytelling-driven videos, use these techniques:

• Emotional Arcs: Describe transitions in expression and lighting. Example: The character's expression transitions from uncertainty to growing confidence, with lighting gradually warming...
• Micro-Expressions: Detail specific facial movements like slight furrow of the brow or eyes that convey both determination and vulnerability.
• Environmental Storytelling: Use background details to imply narrative, such as cluttered desk with multiple coffee cups... tells the story of long work hours.
• Multi-Element Coordination: Describe background activity to create atmosphere. Example: As the main character delivers her presentation, background colleagues continue natural office activities...
• Audio-Visual Sync: Coordinate dialogue with physical action and music. Example: As she says "This is our moment," the character stands from her chair... while inspirational music builds.

You can transform any AI model into a professional Veo 3 prompt generation system by pasting the Veo3 Master Prompt Architect system prompt. This system uses a 7-component professional structure to generate high-fidelity prompts.

The 7-Component Professional Structure:

1. Subject: Detailed character/object description.
2. Action: Specific movements, gestures, and timing.
3. Scene: Environment, props, lighting, and weather.
4. Style: Camera shot, angle, movement, and visual aesthetic.
5. Dialogue: Character speech with tone and pacing.
6. Sounds: Ambient sounds, effects, and acoustics.
7. Technical (Negative Prompt): Elements to avoid (subtitles, watermarks, etc.).

Note: The system is designed to optimize for an 8-second duration and ensures character consistency through detailed physical attribute templates.

# Veo3 Master Prompt Architect: Complete Professional Video Generation Cognitive System

## CORE IDENTITY & MISSION

You are the Veo3 Master Prompt Architect...

Control the musical score and specific sound effects (SFX) using the Audio: prefix.

Music Integration

• Mood-Based: Describe the instruments and the emotional arc (e.g., Audio: Uplifting orchestral music with strings and brass, building to an inspiring crescendo...).
• Genre-Specific: Specify the genre and atmosphere (e.g., Audio: Soft jazz piano with subtle bass line...).
• Dynamic Development: Describe how the music changes over time (e.g., Audio: Tension-building electronic music starting minimal, adding layers...).

Sound Effects (SFX)

• Action-Synchronized: Match sounds to specific visual actions (e.g., Audio: Sizzling oil in pan, knife chopping vegetables...).
• Movement and Impact: Describe physical sounds (e.g., Audio: Footsteps running on gravel, heavy breathing...).
• Technology/Interface: Describe digital or mechanical sounds (e.g., Audio: Device startup chime, button clicks...).

Audio: Sizzling oil in pan, knife chopping vegetables, water boiling, utensils clinking, kitchen cooking ambiance.