How to Use veo3.xyz

A comprehensive guide to creating professional AI video prompts with Veo 3

Getting Started

veo3.xyz is designed to bridge the gap between having an idea and having a usable Veo 3 prompt. The tool guides you through a structured process that ensures you consider every critical element of a professional video prompt.

Veo 3 is powerful, but it responds best to specific, well-structured prompts. Generic descriptions produce generic results. Our tool helps you speak Veo 3's language — the cinematic vocabulary of camera movements, lighting conditions, visual styles, and atmospheric effects that the model was trained on.

This guide will walk you through each field in the generator, explain why each element matters, and show you how to combine them effectively. Follow these principles, and you'll see dramatically better results than with unstructured prompts.

The Three-Step Process

  1. 01

    Describe your scene

    Write at least 2–3 sentences for your scene. Include the environment, the action happening, the time of day, and the emotional tone you are going for. The more specific you are — "a rain-soaked Tokyo alley at 2am" rather than "a street at night" — the better Veo 3 will interpret your intent.

    ❌ A person walking in a city

    ✓ A young woman in a red coat walks alone through a rain-soaked Tokyo alley at 2am, neon signs reflecting on wet pavement, steam rising from vents

    ❌ A beach scene

    ✓ Golden hour light on a deserted beach, gentle waves rolling onto white sand, a lone figure standing at the water's edge watching the sunset

  2. 02

    Set your cinematic parameters

    Choose a camera movement (tracking shot, push in, aerial), a visual style (cinematic, documentary, noir), lighting conditions, and mood. Add subject detail to describe who or what is in the frame. Optionally specify audio — Veo 3 natively generates synchronized sound from your description.

    ❌ Make it look cool

    ✓ Cinematic style, tracking shot following the subject, golden hour lighting, epic mood

    ❌ Add some sound

    ✓ Audio: distant city traffic, rain on pavement, soft jazz from a nearby bar

  3. 03

    Copy, share, and generate

    Hit "Copy Prompt" and paste it directly into Google VideoFX or the Vertex AI Veo 3 API. Use "Share" to get a permalink you can send to collaborators. Your last 5 prompts are saved in session history so you can revisit, tweak, and re-copy without losing your work.

    ❌ Just remember it

    ✓ Copy the prompt and paste directly into VideoFX prompt field

    ❌ Start over each time

    ✓ Use session history to revisit and refine previous prompts

Field-by-Field Guide

Scene Description

Critical

This is the foundation. Spend the most time here. Be specific about environment, action, time of day, and emotional tone. 2-3 sentences minimum.

Subject Detail

High

Describe who or what is in frame. Age, clothing, position, expression. If no human subject, describe the main object or landscape feature.

Camera Movement

High

Choose movement that serves the story. Static for intimacy, tracking for momentum, aerial for scale. Don't add movement just for effect.

Visual Style

Medium

Sets the overall aesthetic. Cinematic for polished narrative, documentary for realism, noir for mystery. Match style to your story.

Lighting

High

Lighting defines mood and time of day. Golden hour for warmth, neon for cyberpunk, harsh sun for tension. Always specify lighting.

Mood

Medium

The emotional atmosphere. Epic, calm, tense, joyful. This guides color grading, pacing, and audio choices.

Audio

High

Veo 3's unique advantage. Describe ambient sounds, specific effects, or dialogue. Audio should reinforce the visual mood.

Aspect Ratio

Low

Match your platform. 16:9 for YouTube, 9:16 for TikTok/Reels, 1:1 for Instagram. This is a technical choice, not creative.

Duration

Low

Longer isn't always better. 5-10 seconds is often ideal for social media. 30 seconds allows for narrative development.

Resolution

Low

Higher resolution takes longer to generate. 1080p is usually sufficient. 4K only if you need it for specific use cases.

Common Mistakes to Avoid

Vague Descriptions

Avoid generic terms like "beautiful," "cool," or "amazing." These don't give Veo 3 specific visual targets. Instead, describe concrete details: lighting, colors, textures, and specific actions.

✓ Fix:

Replace "a beautiful scene" with "a misty mountain lake at dawn, with golden light reflecting off still water and pine silhouettes in the foreground."

Overloading the Prompt

Too many competing visual directions can confuse the model. Focus on 2-3 key visual elements rather than trying to include everything at once.

✓ Fix:

Choose your primary visual goal — is it the lighting? The camera movement? The mood? Build around that core element.

Ignoring Audio

Veo 3's native audio generation is a superpower. Skipping audio descriptions means missing half the immersive experience.

✓ Fix:

Always include audio: ambient sounds, specific sound effects, or dialogue. Audio should complement the visual mood.

Contradictory Elements

Don't combine elements that conflict logically. A "bright sunny day" with "noir lighting" will produce inconsistent results.

✓ Fix:

Ensure all elements of your prompt work together coherently. Lighting, time of day, mood, and style should all align.

Pro Tips for Better Prompts

Start with the Mood

Before writing anything, decide the emotional tone you want. Is this hopeful? Tense? Romantic? Nostalgic? Let the mood guide every other choice — lighting, camera, style, and audio should all serve that emotional goal.

Focus on One Subject

Veo 3 handles single-subject scenes best. If you have multiple subjects, focus on one as primary and keep others in the background or as supporting elements. This creates clearer visual hierarchy.

Consider Pacing

Think about how the scene unfolds over time. Is it a slow, contemplative moment? A fast action sequence? Your camera movement and audio should match the temporal rhythm you want.

Use the Style Dictionary

Our Style Dictionary contains 50+ tested keywords that produce consistent effects. Browse by category and add keywords that match your vision. Each keyword is a proven trigger.

Quick Reference Checklist

Before You Generate

  • Scene description is 2+ sentences with specific details
  • Subject is clearly described (who/what is in frame)
  • Lighting condition is specified (time of day, light source)
  • Mood matches the emotional tone you want
  • Audio description is included (ambient, effects, dialogue)

Quality Checks

  • Prompt is under 800 characters
  • No contradictory elements (e.g., sunny + noir lighting)
  • All elements serve the same mood/story
  • Camera movement makes sense for the action
  • Style keywords complement, don't clutter
Start Creating Prompts