Text to Video Prompt Examples: 40 AI Video Prompts You Can Adapt

Text to video prompts need to describe a moving scene from scratch. Unlike image to video, there is no uploaded frame to preserve. The prompt has to define the subject, scene, camera, motion, lighting, and constraints clearly enough for the model to build a short clip.

Use this guide when you want to start from an idea instead of a still image. If you already have a first frame, use image to video prompts instead.

You can test these examples in Naviya Text to Video and compare model fit in the AI video model comparison.

Text to video prompt formula

Use this template:

Create a [duration] second [format] video.
Subject: [who or what the video is about].
Scene: [location, time, style, lighting].
Camera: [shot size and one camera movement].
Action: [what happens during the clip].
Style: [realistic, cinematic, anime, product, social, documentary].
Constraints: [what to avoid].

The best text to video prompts usually have one subject, one camera idea, and one clear action. Complex stories should be split into multiple short clips.

Cinematic prompt examples

Rainy street hero shot

Create a 6 second cinematic video.
Subject: a lone cyclist wearing a reflective black jacket.
Scene: rainy neon city street at night, wet pavement, blue and violet reflections.
Camera: low angle tracking shot moving beside the bicycle.
Action: the cyclist rides slowly through mist as neon signs flicker in the background.
Style: realistic cinematic lighting, shallow depth of field.
Constraints: no text, no extra riders, no sudden camera cuts.

Desert product journey

Create a 5 second cinematic brand film shot.
Subject: a compact silver travel speaker on desert sand.
Scene: golden hour desert, warm wind, soft dust in the air.
Camera: slow push-in from wide shot to product close-up.
Action: sunlight sweeps across the metal surface while sand moves gently around it.
Style: premium realistic product cinematography.
Constraints: keep the product shape stable, no logo distortion, no people.

Futuristic workspace

Create a 6 second realistic video.
Subject: a designer working at a transparent display in a dark studio.
Scene: modern creative workspace, violet accent lights, soft reflections.
Camera: smooth orbit from behind the shoulder to side profile.
Action: interface lights move across the screen while the designer adjusts a concept image.
Style: clean futuristic production design.
Constraints: no readable private text, no distorted hands, no sudden scene change.

Product video prompt examples

Beauty product ad

Create a 6 second vertical product video.
Subject: a glass skincare serum bottle with a silver cap.
Scene: black acrylic surface, soft violet rim light, mist in the background.
Camera: slow orbit from front-left to center.
Action: light moves across the bottle, condensation appears subtly on the glass.
Style: premium beauty ad, clean and elegant.
Constraints: keep bottle shape stable, no extra products, no unreadable fake text.

Sneaker launch clip

Create a 5 second social product launch video.
Subject: a white futuristic sneaker on a reflective platform.
Scene: dark studio with moving blue and purple light strips.
Camera: low angle push-in.
Action: the platform rotates slowly while the sole catches a bright highlight.
Style: sporty premium campaign.
Constraints: preserve shoe shape, avoid warped laces, no extra shoes.

Food delivery teaser

Create a 6 second vertical ad video.
Subject: a fresh noodle bowl on a dark table.
Scene: warm restaurant lighting, steam rising, shallow depth of field.
Camera: slow overhead-to-front tilt.
Action: steam moves naturally while chopsticks lift noodles once.
Style: appetizing food commercial.
Constraints: keep the bowl stable, avoid messy hands, no text overlays.

For more campaign-specific patterns, use AI video ads prompts.

Social media prompt examples

TikTok-style hook

Create a 5 second vertical social video.
Subject: a creator holding a phone toward camera.
Scene: neon rooftop at night, city lights behind them.
Camera: handheld slow push-in.
Action: the creator turns slightly and the phone screen glows brighter in the first second.
Style: energetic creator content, high contrast, modern.
Constraints: leave safe space for captions, no readable fake app UI, no extra people.

Instagram reel opener

Create a 6 second 9:16 video.
Subject: a desk setup with headphones, notebook, and glowing monitor.
Scene: cozy evening studio, warm desk lamp, violet background light.
Camera: slow pan from left to right.
Action: monitor glow changes softly while paper edges move in a light breeze.
Style: polished creator reel opener.
Constraints: no legible private text, no clutter, keep objects stable.

YouTube Shorts visual loop

Create a seamless 6 second vertical loop.
Subject: a futuristic coffee cup on a desk.
Scene: morning light through blinds, clean workspace, soft steam.
Camera: locked close-up.
Action: steam rises in a repeating pattern, light bands move slowly across the cup.
Style: calm lifestyle short.
Constraints: loop should feel smooth, keep cup shape and position stable.

Anime prompt examples

Rooftop character scene

Create a 6 second anime-style video.
Subject: a silver-haired character in a black jacket.
Scene: city rooftop at sunset, warm clouds, wind, glowing skyline.
Camera: slow push-in from medium shot to close-up.
Action: hair and jacket move in the wind, character blinks once and looks toward camera.
Style: polished anime film look.
Constraints: preserve character design, no sudden outfit change, no extra characters.

Magic object reveal

Create a 5 second anime fantasy video.
Subject: a glowing crystal floating above an old wooden table.
Scene: candlelit room, soft dust, magical blue light.
Camera: slow orbit around the crystal.
Action: light pulses from the crystal while small particles spiral upward.
Style: fantasy anime, clean linework, dramatic lighting.
Constraints: keep the object centered and stable, no scene cut.

Cyberpunk alley

Create a 6 second anime cyberpunk scene.
Subject: a character standing under a neon sign with an umbrella.
Scene: rainy alley, blue and pink signs, wet reflections.
Camera: side tracking shot with background parallax.
Action: rain falls, umbrella tilts slightly, neon reflections ripple.
Style: cinematic anime, moody night lighting.
Constraints: no extra faces, no unreadable text close-up, preserve the character silhouette.

How to improve weak text to video results

If the scene looks generic, add a stronger location, lighting, and camera direction.

If the motion is chaotic, reduce the action to one movement.

If the model invents unwanted objects, add a short negative constraint.

If faces or hands break, choose a closer crop, slower motion, or an image-to-video workflow with a stronger first frame.

If the output ignores the prompt, compare model families before rewriting everything. Some models are better for realism, some for motion, and some for fast exploration.

A practical testing workflow

1. Write one short prompt.
2. Generate one version.
3. Identify the failure mode.
4. Change only one line.
5. Save the winning camera, motion, and constraint phrases.

Example selection workflow

Do not pick a prompt example only because the scene sounds exciting. Pick the example that matches your production problem. If you need product stability, start with a product prompt and keep the object simple. If you need an emotional opener, start with a cinematic prompt and define the subject's action clearly. If you need a social ad, choose a vertical format, a visible hook, and space for captions before adding style.

Create a small prompt library from your own tests. Save the camera phrase, motion phrase, subject constraints, and failure note. A reusable note might say: "slow push-in worked for glass bottle; orbit caused label drift." Over time, this becomes more valuable than a long list of generic prompt examples because it reflects what actually works for your categories.

Try it in Naviya

In Naviya, start with one example from this guide and generate a short clip in the AI video generator. If the subject needs to match a still, switch to image to video. If you are building ad variants, use AI video ads and compare first-second clarity, product stability, and caption space. For deeper motion control, pair these examples with image to video prompts and reverse-engineer AI video motion.

Text to video improves when you treat prompting as production direction. Give the model the shot, the motion, and the limits. Then iterate from the result instead of rewriting the entire scene every time.