You are now in a different world
Welcome 🌟
Table of Contents 🌎
➤ Introduction to Google Veo 3
➤ How to Get Started: Free Subscription
➤ Detailed Guide to the Flow Platform
➤ Overview of the Gemini Video Generation Platform
➤ Audio in Veo 3
➤ Writing Prompts Professionally with ChatGPT
➤ How Can ChatGPT Help You Create a Complete Vlog?
➤ How to Use ChatGPT to Edit Ready-Made Prompts
➤ Can I Write Prompts in Arabic? 🇸🇦
➤ The Best Structure for Veo 3 Prompts
➤ My Method of Writing Prompts (Explained with Examples)
➤ From Idea… to Professional Video: Step-by-Step Scene Execution
➤ How Do I Maintain Character Consistency?
➤ Ready-to-Use Professional Prompt Examples
➤ Common Mistakes and Problems in Prompts
➤ Frequently Asked Questions (FAQ)
➤ Important Legal and Ethical Guidelines
What programs do you need for montage and effects? You will find them at the bottom of this valuable article.
1. Introduction to Google Veo 3
Welcome to the world of AI-powered video creation!
This guide is your gateway to mastering Google Veo 3, Google’s most powerful tool for transforming your ideas into stunning, realistic videos.
Whether you’re a content creator striving to stand out, or simply a curious beginner eager to explore this new world, you’ll find everything you need here—tips, tricks, and hidden gems to unlock your creativity.
In this booklet, you won’t just get theory—you’ll also receive over 10 professional ready-to-use prompts to test right away.
Best of all, you’ll learn how to make ChatGPT your smart creative assistant, helping you refine and customize prompts into unique ideas that carry your personal touch.
🎊 Are you ready to begin your journey into cinematic directing? Let’s go!
2. How to Get Started: Free Subscription
🔑 Step One: Choose Your Plan
Open Google and search for: Google Veo 3.
Select the official link titled: Gemini AI video generator powered by Veo 3.
On the website, go to the Subscriptions section.
Choose the plan labeled: Free Trial Pro (1 Month Free).
🔑 Step Two: Activate Your Subscription
Enter your credit card details.
Don’t worry 👌 — no charges will be made during or after the trial period.
This step is only for verification and activation.
Once activated, you’ll have a Google Veo 3 Pro subscription and automatically gain free access to the Flow platform with 1,000 credits to start experimenting.
🔑 Step Three: Turn Off Auto-Billing (Very Important)
Right after activating your subscription, go to your Google Account Settings.
Disable automatic billing to make sure no charges are applied after the free trial ends.
→ Go to Payments & Subscriptions → Cancel subscription.
By doing this, you’ll ensure they don’t charge you in the future, and your free trial will still remain active until the end of the month. 💯
⚙ 3. Detailed Guide to the Flow Platform
🔑 Important Note:
The Veo 3 model only works with prompts written in English.
If you want to include Arabic dialogue, write the entire prompt in English and insert the Arabic lines inside it — as you’ll see in the examples later.
The Flow platform is the professional interface that gives you full control over every detail of your video.
It’s perfect for anyone who wants precise control over the final look and cinematic output.
🔋 Available Generation Methods
Text-to-Video
Write a detailed text description of your scene, and Veo 3 will transform it into a full video.
Frames-to-Video
Upload a still image, and the model will convert it into a moving video while keeping the same style and environment.
Perfect if you have a shot or preliminary design and want to develop it into a cinematic scene.
Can also be used to extend or continue a scene you’ve already created.
Extend Scene
Available only for ULTRA plan subscribers.
Allows you to increase the duration of generated videos smoothly without breaking the scene.
⚡ Basic Generation Settings
🆗 Golden Tip:
Always start with FAST mode; its results are very close to Quality mode but cost much less in credits.
Veo 3 – Quality: Extremely high quality, but experience shows the difference from Fast is almost negligible.
Veo 3 – Fast: Excellent results when writing a clear and detailed prompt, while saving your credits.
Number of Outputs: Choose 1 to 4 outputs for the same prompt.
Always start with 1 output to avoid spending your credits too quickly. 👉
🆗 4. Overview of the Gemini Video Generation Platform
If Flow is like a “professional studio” with all its tools,
then Gemini is the smart and fast option — perfect for testing ideas and trying prompts before spending your credits.
🔑 How to Access and Use
Go to the website: gemini.google.com
From the homepage, select Video Generation
Paste your prompt into the text box and click Generate
It’s very easy — no complicated settings required
🔑 Important New Feature
You can now upload a still image, and Gemini will turn it into a moving video.
All you need to do is add the image and write a prompt describing the scene or desired motion.
Then, Gemini automatically completes the video for you.
This feature is perfect if you have a shot or design and want to develop it into a cinematic video.
🔑 Why I Recommend Using Gemini
🔰 You get 3 free videos daily (depending on your subscription).
🔰 Perfect for testing your idea and prompt before moving it to Flow.
🔰 Fast results with good accuracy when the description is written clearly.
🔆 My Personal Experience
I always use Gemini as the first step for testing:
I test the camera angles.
I check the character details.
I review the script or dialogue.
If the result is satisfactory, I move the idea directly to Flow to produce a final version with higher quality and full control.
Break the silence with Veo 3 Create high-quality 8-second videos with Veo 3, our latest model designed for AI-powered video creation. You can try it with the Google AI Pro plan or take advantage of the highest level of access with the Ultra plan. Simply describe your ideas, and they'll be brought to life with the built-in voice creation feature.
🔊 5. Audio in Veo 3
🔉 Audio is a key element in any video.
Veo 3 gives you the ability to precisely control audio through your prompts.
🔰 Practical Example:
Try this full prompt in Gemini:
Scene Description:
A calm, cinematic shot of a quiet beach at sunset.
Golden and orange hues reflect softly on the wet sand,
as a small cat walks gently along the shoreline, leaving delicate paw prints behind.
The waves lap gently against the shore, accompanied by natural ambient sounds:
a light breeze passing by, and the crackling of a nearby bonfire adding a warm presence.
The sky transitions seamlessly from pink to purple, then into deep blue
as the sun slowly disappears beyond the horizon.
The flickering firelight casts shifting shadows across the cat’s fur and face,
creating a poetic, dreamlike atmosphere.
Dialogue (spoken in calm tone):
"Everything is silent... except the waves that speak
📝 Tips for Using Prompts
Adding Dialogue:
Write the sentence clearly inside the prompt. Example:
The character says in Arabic: "Hello, beautiful flowers"Adding Sound Effects:
Describe the sounds in detail to make them more realistic. Example:
with realistic ambient sounds of a busy city street, including car horns and people talkingCompletely Silent Scene:
Specify clearly to ensure no sound:
No sound
No ambient sounds
6. Writing Prompts Professionally with ChatGPT
The idea is simple: make ChatGPT your assistant director.
You give it your rough idea, and it transforms it into a detailed, professional prompt that guides Veo 3 to produce a high-quality scene.
Any small idea you have can be turned into a full scenario ready for execution with ChatGPT.
🔘 How to Ask ChatGPT
Don’t settle for a vague phrase like: “Give me a prompt.”
Instead, explain your idea as if you were telling it to a film director.
When framing your request, make sure to include the following elements:
General Idea of the Scene: What happens in the scene?
Character: Appearance, clothing, emotions, and distinctive details.
Dialogue: Exact lines you want, with the tone specified.
Environment: Location, time, weather, and place details.
Technical Requirements: Lighting, camera angle, colors, and any desired or undesired effects.
🔘 Practical Example of a Request
You can ask ChatGPT like this: 👇
Request text in English:
"Hello, I want you to be my director and write a professional prompt in English for the Veo 3 model.
Please turn this scenario into a detailed prompt:"
Scene Prompt (English Version)
Scene:
A selfie vlog of a tiger during golden hour, walking along a dusty mountain road near Ta’if. The atmosphere is warm and peaceful.
Character:
A highly realistic tiger (non-cartoonish) with dusty brown fur and expressive facial features. Wearing a white tank top and a gray winter hat. (Description enhanced creatively.)
Dialogue & Tone:
Speaking in a calm, amazed tone while smiling at the camera:
"Look at this view… it’s like a painting."
Filming & Direction:
Selfie-style camera held by the tiger, with a slight natural shake from walking. Focus on the tiger’s face, showing the road and mountains in the background.
Technical Requirements:
Precise lip-sync with the dialogue.
No text or captions on the video.
Audio limited to the tiger’s sounds and natural ambient noises (light breeze).
7. How ChatGPT Helps You Produce a Complete Vlog
Sometimes the idea is great, but it’s hard to turn it into a well-organized cinematic scene — this is where ChatGPT comes in as your assistant director.
It can break your idea into scenes and turn each scene into a ready-to-use prompt for Veo 3 or Gemini, step by step from concept to the final shot.
👈 Step 1 — Break the Idea into Clear Scenes
Ask ChatGPT to divide your idea into a logical sequence of scenes. For example:
"Write 8 scenes for a vlog featuring a character walking and narrating their adventure on a deserted island."
The output will be a list of short scenes, each with a main idea, objective, and a snippet of dialogue or action — this roadmap makes the next steps easier (filming, shots, transitions).
👈 Step 2 — Turn Each Scene into a Detailed Prompt
Take each scene and ask ChatGPT for a detailed prompt tailored for Veo 3.
Tell ChatGPT to consider artistic and technical elements such as: camera angle, filming style (selfie/POV or fixed camera), dialogue language, lip-sync, and lighting settings.
Example request to send to ChatGPT:
"Turn the following scene into a detailed English-language prompt for the Veo 3 model:"
Scene: (description of location and time)
Character: (appearance, clothing, emotions)
Filming style: selfie POV, camera in character’s hand, slight shake
Dialogue: (dialogue text and tone)
Technical requirements: lip-sync, no text on screen, natural sounds only
👈 Important Tip:
For selfie POV shots to make the scene look natural:
Using a selfie stick:
"POV selfie-style cinematic video, filmed entirely from a selfie stick held by the character…"Directly handheld by the character:
"POV selfie video, shot from the perspective of a handheld camera, held by the character while walking and talking…"
👈 Keep the Character Description Consistent from the Start
One of the most important secrets for continuity in your video is to use a fixed character description in every prompt.
Make sure the appearance, clothing, expressions, and even movement style are written with the same details each time.
Repeat this description without any changes across all scenes,
so the character appears consistent — avoiding confusing changes or distracting differences for the audience.
✋ Practical Example: Desert Vlog — Ready-to-Use Professional Prompt
Scene Type: Hyper-realistic cinematic selfie vlog video in a desert oasis.
Camera Style: Front-facing selfie stick camera held by the character. The camera moves slightly with the natural rhythm of walking on sand.
Time & Setting: Late afternoon in a desert oasis. The golden sun reflects on the calm water surrounded by palm trees. In the distance, sand dunes create a soft backdrop.
Character: A curious desert fox with realistic features — sandy-colored fur, sharp expressive eyes, and natural movements. The fox wears a small traditional vest for a cultural touch.
Action & Dialogue: The fox holds the selfie stick naturally, looking into the camera with playful energy. He speaks entirely in Saudi Arabic with accurate lip sync, saying in a lighthearted tone:
"Look at this view! I didn’t expect to find such a beautiful oasis in the middle of the desert."
Atmosphere & Mood: Warm, adventurous, and full of discovery. The lighting is cinematic, with long shadows from the late afternoon sun.
Audio: Natural ambient sounds only — soft desert wind, distant chirping of birds near the oasis, and the gentle rustle of palm leaves. No music. No extra background sounds.
Important Notes:
• No text on screen.
• No subtitles.
• No captions.
• Photorealistic rendering.
• Arabic only, with natural lip sync.
8. How to Use ChatGPT to Edit Ready-Made Prompts
Sometimes you come across a strong, professionally-built prompt, but its idea doesn’t fit your vision 100%.
Instead of starting from scratch, you can have ChatGPT rework the prompt with the same quality and structure,
while adjusting the content to match your new concept.
✊ Practical Example:
You have the following prompt: 👇
“A photorealistic koala walking upright on the moon, holding a selfie camera, wearing a white beanie.
Shot in handheld vlog style under lunar daylight. Earth appears in the background.”
The result is great, but you want the same style in a different setting.
For example: instead of the sun → Yemen desert at sunset.
All you need to do is write to ChatGPT:
"This prompt is excellent. I want you to keep the same style and structure, but change the concept to:
A koala walking in the Saudi desert at sunset, with the same filming style and realistic details.
Do not change the style — just change the concept and environment."
This way, ChatGPT will return a new version that matches the style but fits your own idea and setting.
👈 Why Is This Method Smart?
It saves you time instead of rewriting the prompt from scratch.
It maintains the high quality that you’ve already tested and know gives excellent results.
It helps you build a series of videos with a consistent style but different ideas.
✍ 9. Can I Write Prompts in Arabic?
With the latest updates, the Gemini platform now supports writing the entire prompt in Arabic,
not just the dialogue ✨
This is a huge difference, because you can write your idea directly, even in your local dialect,
without needing to translate or rephrase it in English.
👈 Important Note:
The Flow platform currently only fully supports prompts in English.
This means you need to write your prompt in English, but you can include dialogue in your own language within it.
🔹 How to Write an Arabic Prompt Correctly
Start by clearly describing the scene (location + time + character).
Specify the camera angle (selfie, outdoor shot, indoor shot, etc.).
Add dialogue in the dialect you want.
Don’t forget the important instructions (e.g., no text on screen, keep audio natural).
Practical Example of a Prompt Written in Arabic
🌐 Full Arabic Prompt Example (Gemini):
Scene: Cinematic shot of a quiet beach at sunset.
Camera: Handheld selfie camera, moving slightly with the character’s steps.
Character: A young Saudi man wearing a light white thobe, smiling and looking at the camera.
Dialogue: Speaks in a calm tone: "Today the weather is peaceful, and the sea cools the heart."
Details: Sun reflecting on the waves, natural sounds of waves and seagulls.
Instructions: No music, no text on the screen.
✋ Example Prompt in English with Dialogue (Flow)
Scene Type: Hyper-realistic cinematic selfie vlog at sunset on a calm beach.
Camera Style: Handheld selfie, moving naturally with the character’s footsteps.
Character: A young Saudi man wearing a light white thobe, smiling and looking into the camera.
Dialogue (in English, precise lip sync):
"Look at this beauty… I didn’t expect to find such a serene oasis in the desert."
Atmosphere & Details: Warm golden light reflecting on the waves, ambient sounds of seagulls and gentle sea breeze.
Instructions: No music, no text or captions on screen.
🏗 10. The Best Structure for a Veo 3 Prompt
⚠ Important Note:
There is no 100% “correct” structure.
Your structure might be different and still work.
But here, we explain the best structure that usually works about 90% of the time.
This structure must be written in English,
with dialogue also in English.
👈 Logical Structure of a Prompt
Scene Description (High-level overview)
Character & Appearance (Detailed look and clothing)
Action & Dialogue (What the character is doing and saying)
Camera & Cinematography (How the scene is shot)
Style & Atmosphere (Mood, lighting, and artistic style)
Technical Constraints (What to avoid, e.g., no text)
Why is this important? 👇
It saves you time — instead of wondering what to write first, the steps are already organized for you.
Produces Strong Prompts → When you include scene, character, dialogue, camera, style, and constraints together, the output is more realistic and professional.
Reduces Errors → Many forget to specify things like “no text on screen” or “type of lighting”, which can lead to unwanted results.
Flexible → You can use the same template for any new idea, just changing the details.
This structure acts like a prompt roadmap that helps you achieve the best results with Veo 3 or Gemini.
Do you want me to show a full example using this template step by step?
👈 Applied Template
Scene Description (High-level overview)
A cinematic vlog-style video set in a quiet desert oasis at sunset.Character & Appearance
A realistic desert fox with sandy-colored fur and expressive eyes. The fox wears a small traditional vest, moving naturally and lifelike.Action & Dialogue
The fox holds a selfie stick and speaks directly to the camera with a playful yet calm tone in Saudi Arabic (accurate lip sync):
"Akhiran laqeit makan mithl hatha… waddi aglis hena lal-abad."
("Finally, I found a place like this… I wish I could stay here forever.")Camera & Cinematography
Front-facing selfie camera, handheld with slight natural sway. Focus on the fox’s face while keeping the oasis and palm trees visible in the background.Style & Atmosphere
Warm cinematic colors with golden sunlight reflecting on the water. The mood feels calm, adventurous, and immersive.Technical Constraints
No background music, no text on screen, only natural ambient sounds (wind, soft bird chirps, rustling palm leaves).
👈 11. My Method for Writing Prompts (Explained Examples)
My method is based on a set of clear and highly effective principles,
which allow me to achieve strong results on the first try without needing multiple attempts.
👀 Core Principles
Reverse-Engineering the Scene
Start from the final image you want, imagining it as a shot from a movie.
Then break it down into elements. For example: 👇
Who is the main character?
What are they doing?
Where are they located?
How do they look and what is their condition/mood?
What is the overall atmosphere and lighting?
Explicit Guidance (Leave No Room for Guessing)
Do not assume the AI will “figure it out by itself.”
Write your instructions clearly and explicitly:
If you don’t want music: No background music.
If you don’t want text on screen: No text on screen.
Layered Details
Start with a very general idea, then add detail layer by layer:
Example:
“A monkey in the jungle”
“A monkey wearing a purple tank top”
“Its fur is shiny and slightly wet”
“Its facial expression reflects curiosity and excitement”
🎬 Be the Director
Treat the prompt as if you are giving instructions to both the actor and the cameraman:
To the actor: “Raise your eyebrow in surprise.”
To the cameraman: “Close-up shot of the face with a blurred background.”
🎨 Use Strong and Precise Descriptions
Instead of “strong” → “massive with well-defined, bulging muscles.”
Instead of “sad” → “his features look exhausted, and his eyes are filled with despair.
🔑 Context is the Key
The more contextual details you add (location, time, overall atmosphere, emotional state),
the more realistic and convincing your result will be.
Example:
Weak: “A man walking.”
Strong: “A tired man walks slowly through a rainy street at midnight, his shoulders hunched under the dim glow of neon lights.”
Perfect example 👌
That’s exactly how you take a simple idea and turn it into a cinematic, production-ready prompt.
Here’s your Mars scene written as a full polished prompt in English (ready to use):
Prompt:
A cinematic, photorealistic scene on the surface of Mars. A curious little girl in a slightly oversized futuristic space suit with a transparent helmet kneels down, her eyes wide with amazement. She slowly reaches out to gently touch a single vibrant red rose emerging from the Martian soil. The camera performs a slow dolly zoom toward the rose, shifting focus softly from the girl’s fascinated face to the flower. The atmosphere conveys innocence and wonder, with surreal beauty in the contrast between the barren Martian landscape and the fragile rose. No sound, no background music, no on-screen text.
.
✅ Example 1: Tiger in the Market (Single Character with Dialogue)
Introduction:
A hyper-realistic selfie vlog video of a tiger exploring a bustling street market in Bangkok, Thailand, during the daytime.
Character Details:
Orange fur with bold black stripes, piercing amber eyes.
Wearing a casual, slightly oversized sleeveless hoodie.
Action:
The tiger is standing near a fruit stall, holding a front-facing camera with one paw while tasting a piece of mango with the other.
The background is filled with people, food carts, and colorful signs.
Tone & Mood:
He speaks in English with a playful, confident tone — mixing curiosity with humor.
Body Language:
Takes a bite of the mango and nods with approval.
Looks around the market with sharp but friendly expressions.
Gestures toward different stalls while addressing the camera.
Dialogue (English):
“Guys, this market is something else! This mango is amazing, but let me show you the food over there.”
Cinematic Style:
Ultra-realistic details, vibrant lighting, and rich colors.
Ambient sounds of a busy market (people chatting, sizzling food, distant motorbikes).
Dynamic selfie POV framing that captures his expressive feline face.
Important Directives:
No on-screen text.
No captions.
No subtitles.
Tiger’s anatomy must stay consistent (realistic tiger body, anthropomorphic expressions).
❗ Why is this prompt successful?
We specified the type of tiger, its location, and detailed characteristics.
✔ We clearly described the tone and dialogue.
✔ We included body language and movements.
✔ We ended with clear technical directives to prevent errors.
✅ Example 2: Bear in the Forest (Single Character with Dialogue) – Copy, Paste, and Try
Scene & Character:
A hyper-realistic cinematic vlog-style scene featuring a large brown bear wandering through a dense green forest.
Bear Appearance:
Thick, slightly messy brown fur with traces of leaves and dust from walking.
Strong, towering build with powerful limbs.
Wearing a rugged backpack strapped across his shoulder.
Facial expression: calm, reflective, with a hint of wonder.
Action & Dialogue:
The bear holds a selfie stick, recording himself as he slowly walks between tall trees.
He stops near a stream, crouches slightly, and lets the water run through his paw before looking at the camera.
Tone: thoughtful, grounded, slightly poetic.
Here’s the translation into American English:
"Everything is silent… except the waves that speak."
Setting & Atmosphere:
A lush forest with tall pine trees, moss-covered rocks, and a gentle stream.
Early morning mist with soft sunlight breaking through the branches.
Natural ambient sounds: rustling leaves, flowing water, distant bird calls.
Camera Style:
Selfie POV with natural shaking from his movement.
Close-up shots of the bear’s face with expressive eyes.
Occasional wide angles capturing the stream and trees around him.
Important Directives:
The bear must remain fully photorealistic.
Anatomy must stay consistent (realistic bear body with subtle anthropomorphic gestures).
No subtitles or on-screen text.
No background music.
✌ 12. From Idea… to Professional Video: How to Transform Your Scene Step by Step
Sometimes you get an amazing idea for a video — a shot, a character, a line, or even just a feeling — but then you get stuck and don’t know how to turn it into a prompt or use the right tool.
Here, we explain the process step by step: from the moment the idea hits you… to the moment you see your video ready in front of you.
👈 Step 1: Write Down Your Idea Immediately
The first rule: if you don’t write the idea down, it will disappear.
Whether you’re in the car, sitting down, or even before bed — open your notes app and jot it down.
Example:
"Selfie vlog — a bear named Dadoob, speaking American English, walking through the Utah desert at sunset, laughing while telling a joke about dinosaurs."
👈 Step 2: Let ChatGPT Expand Your Idea
Go to ChatGPT and have it help you turn your raw idea into a detailed scene.
Ask questions that clarify the picture:
What type of video? (Vlog? Commercial? Horror? Documentary?)
Who is the character? (Bear? Monkey? Robot? Bigfoot?)
What are they doing or saying? (Dialogue? Actions? Facial expressions?)
Where and when is it happening? (Forest? Beach? City? Morning? Night?)
Ask clearly:
"Write me a professional prompt for Veo 3. Scene: selfie vlog. Character walking and talking to the camera. Description: [description]. Location: [description]. Dialogue: [dialogue in the accent]."
Let it add important details like:
Camera angle (selfie handheld? fixed? slight natural shake?)
Lighting (sunset? night neon? full moon?)
Sound effects (footsteps? wind? water? crowd?)
Emotions (playful? scared? calm? sad?)
👈 Step 3: Prepare the Final Prompt
Copy the text ChatGPT gives you.
Edit any details that don’t match your vision.
Keep the character description consistent (clothes, appearance, style) from the start, so it doesn’t change between scenes.
Save your prompts in a file or special notes for easy access later.
👈 Step 4: Choose the Right Platform
Gemini: for quick testing (supports English prompts).
Flow: for high quality and full control (requires English prompts).
👈 Step 5: Review the Result
After the video is generated, ask yourself:
Is the camera angle as I imagined?
Does the character look the same with all the details?
Is the audio clear and the mood right?
If something is missing → edit the prompt and try again.
Here’s the full translation into American English:
Practical Example: From Idea to Video – Copy, Paste, and Try
Scene Type: Hyper-realistic cinematic vlog video in a tropical rainforest.
Character: A large, photorealistic tiger with wet orange fur and bold black stripes. His golden eyes reflect calm focus. He naturally holds a selfie stick camera in his front paw.
Action: The tiger slowly walks through the dense forest during rainfall, then pauses near a small puddle, looking into the camera.
Dialogue: The tiger speaks in clear American English with precise lip sync:
"Even in this rain… peace remains the most beautiful music."
His tone is deep, calm, and poetic.
Camera & Cinematography: POV selfie style, handheld selfie stick, natural shake from walking. Occasional close-ups on the tiger’s face and wide angles showing the wet forest around him.
Atmosphere: Tropical rainforest at dusk, soft gray sky, rain falling gently, vibrant green leaves glistening with raindrops.
Audio: Natural ambient sounds only — rain, rustling leaves, distant tropical birds. No music.
Constraints: No subtitles, no text, no artificial overlays. The tiger must remain fully photorealistic with consistent anatomy.
✔ Golden Rule: ☝
AI doesn’t read your intentions… it follows your description literally.
The more precise and clear your description, the closer the video will be to your original idea.
💂 13. How to Keep Your Character Consistent
One of the biggest challenges when using Veo 3 is keeping your character looking the same across multiple videos.
The first clip might come out perfect, but when you make a second video… you might be surprised to see a completely different character 😅 (different face, changed color, or even different clothing).
❗ Why Does This Happen?
The Veo 3 model doesn’t have a “memory” of previous videos.
Each time, it generates the result solely based on the text description you provide.
Any small change in the description is interpreted as a new character.
Even if you repeat the same description 100%, the result usually isn’t perfectly identical.
The similarity typically reaches around 80%-90% — which is good, but not complete.
👈 Solution: How to Keep Your Character Consistent
Lock the description 100%.
Do not change a single word in the character’s description.👈 Example: 👇
"A realistic chimpanzee with dark, expressive fur. He wears a fitted blue denim jacket and stylish sunglasses pushed up on his forehead."
Use Distinct Visual Elements
Add details that never change:
Rare fur color (golden, silver).
A consistent accessory (glasses, hat, bag).
A clear logo or pattern on the clothing.
Maintain Continuity
If you’re creating a series of scenes, don’t let the character disappear and reappear suddenly.
Keep it moving sequentially, as if it’s the same being in the same world.
👈 Golden Tip
Save your character description in a notes app or a dedicated file.
Copy the exact same description for every prompt.
Only change the location, time, or camera angle.
But never touch the character’s appearance description.
This way, even if there are minor differences, viewers will recognize it as the same character without confusion.
14. Ready Professional Example 👉
👈 The Cat in New York
Scene Type: Hyper-realistic front-camera selfie vlog.
Camera Style & Movement:
Front-facing selfie POV, naturally held by the cat (no phone visible). Slight handheld shake from walking. The camera stays focused on the cat’s face while glimpses of the busy streets appear in the background.
Time & Setting:
Sunset in New York City. Tall skyscrapers, neon lights starting to turn on, yellow taxis passing by quickly, and distant honking clearly audible.
Character:
Realistic-sized brown-orange tabby cat standing in a humanoid posture.
Wide, expressive eyes reflecting the city lights.
Wearing a black hoodie with the NYC logo on the chest.
Small bag slung over the shoulder.
Friendly yet smart expression, confident in front of the camera.
Voice & Lip Sync:
Perfect lip-sync with a youthful male voice, clear American accent, speaking English:
"Man, this city never sleeps… but I kinda like it. Heh."
Rendering Style:
Cinematic photorealism: detailed fur, neon light reflections on eyes and clothing, warm sunset light blended with cool street lighting.
Directives:
No on-screen text.
No subtitles.
No logos or interface elements.
—----------------------------------------------------------------------------------------------------------------
👈 The Challenging Giraffe 👇
Scene Type: Hyper-realistic selfie-stick style vlog video.
Camera Style & Movement:
Front-facing camera held by the giraffe using a selfie stick. Natural motion as the giraffe walks slowly and assertively toward the camera.
Time & Setting:
Daytime in a dry savannah-like area in Kenya. Strong natural sunlight casts clear shadows.
Character:
Realistic, tall, muscular giraffe with detailed patterned fur and expressive eyes.
Long neck slightly bent forward, head held high with intense eye contact.
Facial expression: assertive and confident.
Body Language:
Moves forward with deliberate steps.
Tilts head slightly toward the camera in challenge.
Uses neck gestures to emphasize presence.
Voice & Lip Sync:
Perfectly synced to a deep, commanding Saudi Arabic voice:
"Bottom line… any real men among you humans, come to me here in Kenya! Got a pimple on your head? I’ll crush it for you."
Mood:
Highly assertive, challenging, and confident.
Directives:
No subtitles.
No on-screen text.
No background music
14. Ready Professional Prompt Examples (Continued)
👈 The man in space
Scene Type: Hyper-realistic cinematic video inside a spacecraft.
Camera Style & Movement:
Static camera placed at a distance inside the spacecraft, showing the entire body of the bear. Slight camera shake for realism.
Time & Setting:
Inside a dimly lit, realistic space capsule. Emergency red lights flash slowly, casting dramatic shadows.
Character:
A large, realistic brown bear wearing a miniature astronaut suit.
Visor lifted, showing expressive face full of panic.
Positioned near a control panel, reacting to the emergency.
Voice & Lip Sync:
The bear shouts in a realistic Saudi Arabic male voice, perfectly synced with mouth movements:
"Guys! Guuuys! The oxygen is running out! What do we do? We’re gonna die!"
Mood & Expression:
High stress and urgency. Loud voice, wide eyes, frantic paw movements, full-body tension.
Directives:
No subtitles.
No on-screen text.
No music.
—---------------------------------------------------------------------------------------------------
🔹 The Sheep’s Vlog in the City
Scene Type: Hyper-realistic cinematic selfie-style travel vlog video, filmed in 4K during daytime.
Setting:
A vibrant, modern city under natural daylight. Background features skyscrapers, vehicles, and pedestrians. Sunlight reflects off buildings, creating lens flares and sharp contrast.
Camera Style & Movement:
Front-facing camera held steadily by the blonde sheep in selfie mode, swaying naturally as it walks. Focus shifts between its face and the dynamic city backdrop.
Character:
A fully realistic blonde sheep with detailed fur, expressive eyes, and natural facial expressions.
Wearing a fitted open blue denim jacket, stylish sunglasses resting on the forehead.
Posture: confident yet casual.
Body Language:
Gestures naturally with free hoof while speaking.
Dialogue (--------):
"See how you can create amazing content with AI… but keep it a secret!"
Audio:
Natural ambient city noise (traffic, footsteps, wind).
Directives:
No subtitles.
No on-screen text.
No music.
Copy and paste
👈 Donkey in Space – Technical Version
Scene Type: Hyper-realistic cinematic video inside a spacecraft.
Camera Style & Movement:
Static camera placed at a distance inside the spacecraft, showing the entire body of the donkey. Slight camera shake adds realism to the scene.
Time & Setting:
Inside a dimly lit, realistic space capsule. Emergency red lights flash intermittently, casting shadows across the cabin walls.
Character:
A realistic donkey wearing a miniature astronaut suit.
Visor lifted, showing a face full of focus and stress.
Interacting urgently with the control panel.
Voice & Lip Sync:
The donkey speaks in a calm but urgent Saudi Arabic male voice, perfectly synced:
"The situation is critical… Oxygen at 15%, capsule pressure 0.8 bar, we need to override the alarm before systems fail!"
Mood & Expression:
High tension, extreme focus and worry. Wide eyes, sharp and frantic hoof movements.
Audio:
Warning alarms and sirens inside the capsule
Mechanical hum of devices and oxygen system
No music, only ambient emergency sounds to heighten pressure
Directives:
No subtitles
No on-screen text
No music beyond emergency sounds
—-------------------------------------------------------------------------------------------------------------------
👈 The cat's vlog in the city
Scene Type: Hyper-realistic cinematic selfie-style travel vlog video, filmed in 4K during daytime.
Setting:
A vibrant, modern city under natural daylight. Background features skyscrapers, vehicles, and pedestrians. Sunlight reflects off buildings, creating lens flares and sharp contrast.
Camera Style & Movement:
Front-facing camera held steadily by the cat in selfie mode, swaying naturally as it walks. Focus shifts between its face and the dynamic city backdrop.
Character:
A fully realistic brown-orange tabby cat with textured fur and expressive eyes.
Facial features realistic and slightly anthropomorphic.
Outfit: fitted miniature blue denim jacket, stylish sunglasses on forehead.
Posture: confident yet casual, walking upright like a humanoid.
Body Language:
Gestures naturally with its free paw while speaking.
Dialogue (American English:
):
"Today I taught you how to prepare a prompt and make a professional video with AI!"
Audio:
Natural ambient city noise (traffic, footsteps, wind).
Directives:
No subtitles
No text
No music
💪😅تكملة
🔹 Chimpanzee in the forest
Style Guide: Photorealistic cinematic style.
Scene Type: Natural vlog-style documentary scene.
Camera: Handheld, shoulder-level, realistic shake.
Setting: Dense, humid forest in the Pacific Northwest (Washington/Oregon).
Tall pine and fir trees.
Thick moss on trunks and rocks.
Narrow dirt trail with scattered foliage.
Sunlight filters softly through branches.
Character: Chimpanzee — fully realistic, dark brown fur with natural texture.
Movements: Agile, natural walking on two legs or slightly hunched posture.
Facial Expression: Expressive and curious.
Action & Dialogue:
The chimpanzee walks forward through the trail.
Turns its head slightly toward the camera, speaking casually in American English:
"Hey guys… I hope the tutorials were helpful, but most importantly: apply what you’ve learned."
Cinematic Details:
Late afternoon lighting, soft golden tones.
Long, realistic shadows.
Natural ambient sound: wind, bird calls, tree creaks.
No music, no text, no overlays.
—---------------------------------------------------------------------------------------------------
🔹 Sad owl
Style Guide: Photorealistic cinematic style.
Scene Type: Interview-style shot.
Camera: Static, frontal, eye-level.
Setting: Cold snowy Himalayas (high-altitude).
Snow-covered rocks, pine trees, fog.
Light snowfall, visible frost.
Character: Owl — fully realistic, large, with detailed feathers in shades of white and gray.
Face Expression: Calm and slightly sad, big expressive eyes.
Body Language: Minimal movement, slight shifts of wings or head.
Dialogue (American English:):
"Guys… I feel a bit wronged. I feel like people like the monkey character more than me, even though he and I are so similar hahaha."
Cinematic Details:
Cold blue lighting, realistic shadows.
Subtle snow falling.
Ambient audio only: cold wind, distant forest sounds.
No music.
No text or overlays.
👈 The talking tree
Style Guide: Photorealistic cinematic style.
Scene Type: Interview-style vlog.
Camera: Shoulder-level, static and steady.
Setting: Dimly lit forest clearing.
Tall surrounding trees, scattered leaves, roots visible.
Subtle mist in the air.
Character: Realistic ancient tree.
Textured bark, moss patches, natural knots and branches.
Thick, gnarled trunk with expressive “face” in the bark.
Branches subtly move as if gesturing.
Body Language: Minimal but natural movements (slight sway, leaves rustling).
Dialogue (American English):
"I was here before humans even thought of AI… and are you sure you understand me?"
Cinematic Details:
Natural soft sunlight filtering through canopy.
Shadows cast realistically on ground.
Ambient forest sounds only (wind in leaves, distant birds).
No music.
No text or overlays.
👇👇👇--------------------------------------------------👇👇👇
🔹 Darth Vader - Empire Vlog
Style Guide: Photorealistic cinematic vlog style.
Scene Type: Behind-the-scenes vlog.
Camera: Handheld front-facing, slight shake.
Setting: Dark metallic corridor (Death Star / Imperial Star Destroyer).
- Dim moody lighting with subtle flicker.
- Ambient: mechanical hums, air circulation, footsteps.
Character: Darth Vader.
- Armor realistic, heavy, with natural reflections.
- Helmet slightly worn with battle scuffs.
- Breathing sound present but subtle.
Body Language: Faces camera, slight head tilt.
Tone: Calm, deep, with frustration.
Dialogue (American English:):
"Imagine… even I need a break. This laser can’t even cut an apple anymore. And the empire? It’s all drama."
Cinematic Details:
- Real handheld micro-shake.
- Shadows move across armor naturally.
- No music, no effects, no overlays.
✨ 14. Ready-made professional prompt examples (continuation)
🔹 Stormtrooper - Drama Backstage
Style Guide: Photorealistic cinematic vlog style.
Scene Type: Behind-the-scenes vlog.
Camera: Handheld, front-facing, slight shake.
Setting: Imperial base / Death Star corridor.
- Dim industrial lighting, slight flicker.
- Ambient: machines, air vents, footsteps.
- Background: passing troopers or empty chairs.
Character: Stormtrooper.
- Armor slightly scratched and dusty.
- Body posture: relaxed, slouched.
- Voice: annoyed, sarcastic.
Body Language: Holds camera at chest level, angled up.
Breathes with frustration, shakes head.
Dialogue (American English:):
"Ah man… I’m tired of acting in this video. Please, give me a break!"
Cinematic Details:
- Realistic armor reflections and scratches.
- Subtle helmet shadowing.
- Sync with Arabic voice.
- No music, no text, no overlays.
---
🔹 Modern Day Ninja
Style Guide: Photorealistic cinematic vlog style.
Scene Type: Casual selfie-style vlog clip.
Camera: Handheld, front-facing, slight shake.
Setting: Traditional Japanese ninja dojo in the forest.
- Wooden structures, bamboo, stone paths.
- Early morning or soft afternoon light.
- Ambient: birds, wind, rustling leaves.
Character: Modern ninja.
- Dark tactical ninja gear, worn fabric, dusty.
- Face partially visible under hood or mask.
- Posture: relaxed but alert.
Body Language: Holds camera low, tilts head while speaking.
Occasional shrug or sigh.
Dialogue (American English:):
"Guys… today I’m training in ninjutsu. Looks like I’m about to clone myself. Send your prayers!"
Cinematic Details:
- Natural sunlight with moving shadows.
- Fabric moves realistically.
- No music, just ambient sounds.
- No overlays, no filters.
✨ 14. Ready-made professional prompt examples (continuation)
🔹 The student in the lecture hall
Style Guide: Photorealistic cinematic vlog style.
Scene Type: Casual university vlog.
Camera: Handheld selfie, slightly shaky.
Setting: Large modern university lecture hall.
- Rows of desks, chairs, projector screen.
- Students chatting in the background.
- Daytime with natural light through windows.
Character: Young Saudi male student.
- Wears hoodie and backpack.
- Messy hair, tired expression.
Body Language: Adjusts camera nervously, shrugs, sighs.
Dialogue American English:):
"Honestly, I’m tired of classes… the professor is explaining like I’m on another planet."
Cinematic Details:
- Photorealistic lighting and textures.
- Subtle crowd noise, pens tapping, faint chatter.
- No text, no overlays, no music.
🔹 The medieval knight Brumbit
Style Guide: Photorealistic cinematic vlog.
Scene Type: Medieval knight monologue.
Camera: Static, tripod-like, steady shot.
Setting: Open medieval battlefield at dawn.
- Rolling fog, grass fields, distant burning village.
- Horses and soldiers faintly visible in the background.
Character: Medieval knight.
- Realistic steel armor with dents and scratches.
- Heavy sword resting on shoulder.
- Dirty, sweaty, tired face under helmet.
Body Language: Leans on sword, breathes heavily, stares into camera.
Dialogue (American English):
"We fought all night… but what’s the point of blood if the war never ends?"
Cinematic Details:
- Morning fog, realistic breath condensation.
- Armor reflections and dirt detail.
- No text, no overlays, no music.
✨ 14. Ready-made professional prompt examples (continuation)
🔹 The giant dragon
Style Guide: Photorealistic cinematic vlog.
Scene Type: Epic fantasy vlog.
Camera: Aerial wide-angle shot, sweeping motion.
Setting: High snowy mountain peaks.
- Storm clouds swirling.
- Lightning in the distance.
Character: Giant dragon.
- Massive wings spread wide.
- Dark metallic scales glinting in lightning.
- Glowing orange eyes, sharp teeth.
Body Language: Roars into camera, then lowers head to speak.
Dialogue (American English):
"I am the king of the mountains...and whoever challenges me will be reduced to ashes."
Cinematic Details:
- Lightning flashes reflecting off scales.
- Snowstorm winds blowing through scene.
- Ambient: thunder, heavy wind.
- No overlays, no text.
🔹 The dancing rabbit
Style Guide: Photorealistic cinematic vlog.
Scene Type: Fun, dance-style vlog.
Camera: Handheld selfie, shaky with rhythmic movement.
Setting: Urban rooftop at sunset.
- Graffiti walls, distant city skyline.
- Warm orange and pink sky.
Character: Realistic anthropomorphic rabbit.
- White fur with gray patches.
- Wearing a hoodie and sneakers.
- Energetic, cheerful expression.
Body Language: Jumps, spins, dances to imaginary beat.
Dialogue (American English):
"Let’s gooo… time to rock the stadium… the rabbit has arrived!"
Cinematic Details:
- Realistic fur and clothing movement.
- Subtle city sounds: traffic, wind.
- No text, no overlays, no actual music.
✨ 14. Ready-made professional prompt examples (continuation)
🔹 The loyal dog
Style Guide: Photorealistic cinematic vlog.
Scene Type: Emotional vlog.
Camera: Handheld selfie, slightly shaky, close-up on face.
Setting: Quiet suburban street at dusk.
- Street lamps turning on, orange sky.
- Parked cars and trees in background.
Character: Realistic dog (German Shepherd).
- Detailed fur with black and tan patterns.
- Sad but loyal eyes.
Body Language: Holds camera close, sighs, ears droop.
Dialogue (American English):
"Even if you leave me… I’ll always be waiting for you."
Cinematic Details:
- Subtle breath visible in cool air.
- Ambient: faint cars, crickets.
- No overlays, no text, no music.
🔹 The adventurous duck
Style Guide: Photorealistic cinematic vlog.
Scene Type: Comedy travel vlog.
Camera: Handheld selfie stick, swaying naturally.
Setting: Busy city canal (like Amsterdam).
- Boats, bridges, people walking.
- Daytime with clear skies.
Character: Realistic duck.
- Yellow feathers, small backpack on back.
- Excited, adventurous tone.
Body Language: Waddles while filming, flaps wings playfully.
Dialogue (American English:):
"My journey started here… and what’s coming is even greater… kids, fasten your seatbelts!"
Cinematic Details:
- Realistic water reflections.
- Ambient: ducks quacking, water splashes, city chatter.
- No overlays, no captions.
—----------------------------------------------------------------------------------------------------------
🔹 Space Rabbit
Style Guide: Photorealistic cinematic vlog.
Scene Type: Sci-fi comedic vlog.
Camera: Handheld selfie, slightly shaky.
Setting: Inside futuristic alien spaceship.
- Holographic panels, glowing lights.
- Outer space visible through window.
Character: Realistic anthropomorphic rabbit.
- White fur with glowing blue patterns.
- Wearing a small space suit.
Body Language: Floats slightly in zero gravity, laughs, makes goofy faces.
Dialogue (American English:):
"😂… Imagine a rabbit piloting a spaceship! And I’m faster than light!"
Cinematic Details:
- Glow effects on fur patterns.
- Ambient: spaceship hum, faint beeps.
- No overlays, no text.
🛠 15. Common Prompt Issues and Mistakes
Even with experience… mistakes can happen.
Here’s a collection of the most repeated issues I and others have encountered, along with their solutions:
❌ Issue 1: Character looks different from scene to scene
✅ Solution: Copy the exact same description in every prompt; don’t change a single word.
❌ Issue 2: Strange text or logo appears on screen
✅ Solution: Always include in your prompt:
"No text, no captions, no subtitles, no logos."
❌ Issue 3: Camera isn’t as you imagined (different angle)
✅ Solution: Specify the camera clearly, for example:
"POV selfie-style video, front-facing handheld camera with natural shake."
❌ Issue 4: Audio is off or not synced
✅ Solution: Add in the prompt:
"Perfect Arabic lip sync" + specify the accent.
❌ Issue 5: Music is added without request
✅ Solution: Always write: "No background music."
❌ Issue 6: Scene looks dark or unclear
✅ Solution: Specify lighting clearly, such as: "Bright daylight" or "Golden hour sunlight."
❓ FAQ – Frequently Asked Questions
Q: Do I have to write in English?
A: In Gemini, you can write fully in Arabic, but Flow requires English (except for dialogue).
Q: Is there a way to keep the same character consistent?
A: The only solution is to repeat the exact same description without any changes.
Q: Can I create a long video (minutes)?
A: Currently, no. The maximum is less than one minute. But you can create short scenes and combine them in editing.
⚖ 16. Important Legal and Ethical Guidelines
• Do not use Veo 3 or Gemini to create content that harms or disrespects real people.
• Do not impersonate someone in a harmful way.
• Respect intellectual property — do not use copyrighted characters without permission.
💡 In Conclusion:
This booklet is just the beginning.
The key is in your hands: experimentation, creativity, and repetition.
Every prompt is a new opportunity to discover a different style.
✨ Enjoy creating your own worlds… and be the director.
Download CapCut for free from the Google Play Store: Go
Note: In this lesson, we learned how to use commands consistently. You can start experimenting now, but keep in mind that the free plan will expire soon. You can try other free plans or alternative apps, but this platform delivers highly accurate results.
If you like the platform, consider subscribing to a paid plan to continue pursuing your passion and taking your skills to the next level. However, please follow ethical guidelines: the platform is not responsible for any inappropriate content, such as nudity or harmful material.
Create professional content that generates income safely, stays clear of ethical concerns, and helps you succeed responsibly.
AI video generator
create AI videos free
best AI video tools
text to video AI
AI video editing software
AI content creation tools
how to make videos with AI
free AI video maker online
realistic AI video generator
AI animation video maker