Veo 3 is an AI model for creating cinematic 8-second videos directly on Telegram

Example prompt: A woman wearing a beige blazer. On the screen behind her, it says “VEO3 on @GPT4Telegrambot.” A news broadcast setting, an American anchorwoman is sitting at a desk, speaking in clear English: “Let’s figure out how Veo 3 works and how to use it to create high-quality videos.”

Two ways to generate:

By text description — simply write what you want to see
Based on an image — upload a photo and describe what should happen in it

How to Create a Video?

1️⃣ Go to @GPT4Telegrambot
2️⃣ Enter the command /video → Veo 3
3️⃣ Fill in the parameters in the menu that opens:

1. Prompt

Describe what should happen in the video. Use English. The more precise and visually detailed your description is, the better the result.

Example prompt: An elderly woman wearing an astronaut suit, speaking in clear English: “Retirement is so boring. I read Elon Musk’s tweet and now I want to fly to Mars.”

2. Image

Upload a photo that you want to:

Animate
Use as the first frame of the video
Add as a visual reference

Example prompt: A man says, “You can create any video from a photo using Veo 3 in the largest AI bot on Telegram. It’s easy!”

3. Duration

The video always lasts exactly 8 seconds. If you set a scene or dialogue that’s too long, it may get cut off. Use every second wisely.

4. Veo 3 Versions

The bot offers two generation options:

Veo 3 Fast — faster and cheaper. Generally follows the prompt accurately, with realistic physics and sound effects. Perfect for tests and drafts.
Veo 3 — produces highly realistic videos with sound effects and speech. Best for physics, camera work, and atmosphere. Uses two generations.

Aspect Ratio

Veo 3 generates videos in 16:9. If you use an image with a different aspect ratio, the final video will still be 16:9 — with black bars on the sides to preserve proportions.

After you set the parameters, click “Start Generation.”

⏳ Generation takes about 5 minutes.

Mastering Veo 3

Want the character to speak?

Don’t include actions in the same sentence. Write it briefly and clearly: “Speaks in clear English.” If you add something like “jumps and speaks,” the model may focus on the action and ignore the speech.

Example prompt: A rocker-grandma livestream. She sits in front of the camera and says: “I can say anything you write for me in the prompt in the largest AI bot on Telegram!”

You can set an accent for characters. Simply specify the desired accent in the prompt, and the AI model will voice the character accordingly.

Want to keep the same character across different videos?

Write a detailed description or use the same image. For example, take a screenshot of the last frame from a previous video and use it as the base for the next one — this helps preserve the character and overall atmosphere.

Example prompt: A grandmother speaking in clear English with a bright, lively voice: “To continue the video, take a screenshot of the last frame and use it as the base.”

Specify everything that matters

You can specify everything down to the smallest details: music, background sounds, accents, emotions. The prompt can be long — the key is making sure everything fits within 8 seconds.

Example prompt: The video is a continuous, cinematic one-shot from the point of view of a free-spirited lion exploring a vibrant green, abandoned city. The camera follows the natural movement of a little lion. The journey begins through a bustling ivy-covered gate, zig-zagging between a well, beams, and bricks. The lion cub trips and falls. It bursts onto a balcony overlooking the sea, skids past stairs, and startles birds. A sudden turn reveals a huge balcony overlooking the blue ocean. [Describe movement and sensory details here] as the character continues forward. Suddenly, the scene opens up to reveal a breathtaking view: the lion cub approaches a beautiful chair on the balcony, sits on it, and looks out at the stunning nature. The character experiences delight, soaking in the atmosphere. Visual style is hyper-detailed with tactile textures like castles, trees, ocean. Camera movement: [Describe POV or camera type] with [describe camera behavior — fast cuts, smooth glides]. Mimics [subject’s movement style or personality]. Main subject: [the journey of a little lion cub]. Background setting: [general geography or cultural setting] with elements like [markets, natural landscapes, architecture, etc.]. No explicit signage — everything is conveyed through visual storytelling techniques. Lighting / Mood: [describe how lighting changes through the journey]. The overall mood is [emotional tone — playful, intense, dreamy, etc.].

Example prompt: A cinematic low-angle front shot captures a sleek, high-performance sports car accelerating through a dimly lit urban street at night. The camera is positioned just inches from the asphalt, tracking fast and close to the front bumper as the car roars forward with aggressive speed. Sparks fly as the tires screech into a sudden drift, the headlights casting sharp beams across the wet pavement and nearby buildings. Neon signs and streetlights blur in the background, enhancing the sense of velocity. The camera shakes slightly with handheld intensity, maintaining focus on the car's grille and front wheels as it powers through the corner. The lighting is high contrast with cool tones and glowing highlights, delivering a gritty, high-octane cinematic look.

How to Write an Effective Prompt?

Example prompt: A young woman drives at high speed, gripping the steering wheel, focused and tense. In the passenger seat, a large kangaroo casually eats a bag of Lay’s chips, holding it with both paws and munching in a very human-like way. The car interior shakes from the speed, and through the windows we see a blurred landscape — extreme velocity. The woman, irritated, glances at the kangaroo and says in clear English: “They’re chasing us, and I want chips too!” The kangaroo keeps eating, completely unconcerned.

It’s important to understand: you only have 8 seconds to fit in the storyline, movement, atmosphere, dialogue, and effects. Your prompt should be precise, logical, and detailed.

1. Start with the main point: Who is doing what?

Specify the character: man, woman, child, cat, robot, etc.
What are they doing? Use one clear verb: walking, speaking, looking, waving, sitting, drawing.

❌ Bad: “A lady is walking on the beach, singing a song, smiling, dancing, and talking.”
✅ Good: “A lady is slowly walking on the beach, speaking in clear Spanish.”

2. Describe the setting (if it matters)

Where is the action: in a desert, in a big city, by the sea, in space. What does the scene look like: lighting, colors, style.

Example: “A dark room with purple neon lights, the camera slowly zooms in.”

Example prompt: The video begins with a person walking through snow; the camera is on their boots. Tense music and the crunch of footsteps in a blizzard. The camera slowly pulls back as the person walks into the mountains, then they stop and say in clear English: “Try Veo 3 now.”

3. Sound effects

You can specify:

Atmosphere: “tense music playing,” “sound of footsteps in the snow.”
Effects: “sound of wind,” “object falling,” “city background noise.”

4. Define the style and mood

Do you want the video to look cinematic, futuristic, or like an animated film? Use words such as: cinematic, realistic, surreal, 3D render, three-dimensional visualization.

Now it’s your turn to write a prompt and see what Veo 3 can do. Good luck!

@GPT4Telegrambot