Prompt structure
[Shot and camera] + [Hero] + [Action] + [Place] + [Style and light] + [Sound]
Describe sound in three layers:
- Dialogue: in quotation marks. The man says: «Try it yourself».
- Sound effects: be specific. Thunder rumbles in the distance.
- Ambience: separately. Ambient: quiet café chatter.
The one main sound rule: one line, one main effect and one background work better than a sound mess. If you do not need music, write no music.
Ads and products
1. Product close-up
Slow push in on a matte black thermo mug on a wooden table, steam rises from the open lid, soft morning light from the window, shallow depth of field. Sound: quiet morning ambience, a spoon clinks on ceramic. No music.
2. Unboxing
Top-down shot, hands open a craft paper box on a white bed and lift out a knitted scarf, natural daylight. Sound: paper rustling, soft fabric sounds. No music.
3. Coffee in a café
Close-up, milk is poured into a cup of espresso forming latte art, slow motion, warm café light. Sound: espresso machine hiss, quiet café chatter in the background.
4. Cosmetics
Macro shot, a drop of serum falls from a glass pipette onto a clean surface, light refracts through the drop, soft pastel background. Sound: a single soft drip. No music.
People and speech
The lines in these prompts are in Russian on purpose, because the heroes speak Russian. Replace them with lines in the language your hero should speak.
5. Talking to the camera
Medium shot, a young woman in a cozy kitchen looks into the camera and says: «Я больше не трачу на это вечер». Natural daylight, static camera. Ambient: quiet kitchen sounds. No music.
«Я больше не трачу на это вечер» means «I no longer spend an evening on this».
6. Expert
Medium close-up, a man in a grey sweater sits in a bright office, looks into the camera and says: «Начните с одной простой задачи». Soft window light, slow push in. No music.
«Начните с одной простой задачи» means «Start with one simple task».
7. Dialogue
Two friends sit at a café table. The woman says: «Ты это сам сделал?». The man smiles and says: «За один вечер». Static camera, warm light. Ambient: café chatter.
«Ты это сам сделал?» means «Did you make this yourself?», «За один вечер» means «In one evening».
8. Greeting
Medium shot, a woman raises a glass in front of a decorated Christmas tree and says: «С Новым годом!». Warm string lights, slow push in. Sound: glasses clink, soft festive music.
«С Новым годом!» means «Happy New Year!».
Lip sync in Russian is weaker than in English: keep phrases short. More about talking heroes: AI talking avatar.
Atmosphere and nature
9. Night city
Wide shot, a rainy city street at night, neon signs reflect in puddles, a person with an umbrella walks past, static camera. Sound: rain, distant traffic. No music.
10. Forest
Slow push in through a misty pine forest at sunrise, sun rays break through the trees. Sound: birdsong, light wind. No music.
11. Sea
Low angle shot, waves crash on dark rocks at sunset, spray glows in the warm light, slow motion. Sound: waves and wind.
12. Thunderstorm
Wide shot, dark storm clouds over a wheat field, lightning flashes on the horizon, wind moves the wheat. Sound: thunder rumbles in the distance, strong wind.
Story shots
13. The start of a story
Tracking shot, a girl with a red backpack runs to catch a departing tram on a sunny autumn street, leaves fly under her feet. Sound: tram bell, footsteps. No music.
14. Cozy ending
Medium shot, an old man reads a book in an armchair by a fireplace, a cat sleeps on his lap, slow push in. Sound: crackling fire, page turns.
15. From a photo (image-to-video)
The person in the photo slowly turns toward the camera and smiles, a light breeze moves the hair. Static camera, the face stays the same. Ambient: quiet outdoor sounds. No music.
Which version to choose
| Version | When to use | 6 s, 1080p, with sound |
|---|---|---|
| Veo 3.1 Lite | drafts, simple scenes | ~84 rub. |
| Veo 3.1 Fast | good quality at a low price | ~126 rub. |
| Veo 3.1 | final, close-ups, maximum detail | ~417 rub. |
Prices on KADR (kadr-ai.ru), October 2026. This is not a public offer.
If you do not need sound, switch it off: on the top Veo 3.1 that is half the price, on Fast and Lite you save 20-40%. More: how to add voice to a video with AI. Model review: Veo 3.1. Prompts for other models: Kling 3.0, Seedance 2.0.
In short
A Veo 3.1 prompt: shot and camera, hero, action, place, style and sound. Lines in quotation marks, effects and ambience separately, «no music» if you do not need music. A clip is 4, 6 or 8 seconds. Make drafts on Lite and the final on Fast or the top Veo 3.1.
