Skip to main content

Which AI Video Model to Use

Pick the right AI video model for your clip

Khachatur Gharibyan avatar
Written by Khachatur Gharibyan
Updated over 2 weeks ago

Model

When to Use

Runway Gen-4 Turbo

When you have an image you want to animate into a quick, stylized video loop or concept clip.

Veo 3 (Google)

When you need a full-length video from text with high realism and built-in audio (voice, effects, music).

Seedreams 1.0 Pro

When you want super-realistic videos from text at a lower cost than Veo 3, balancing quality with efficiency.

Seedreams 1.0 Lite

When you need fast, low-cost video drafts or simple text-to-video experiments and can accept occasional glitches.

Hailou 02

When you need a cinematic video clip with smooth, realistic motion and strong object consistency.

Hailou 01-Director

When you need a shot with precise camera moves and framing—ideal for teaser or trailer highlights.

Seedance 1.0 pro prompt guide

Seedance1.0 pro has a strong ability to follow semantic instructions. By writing good prompts, it can generate rich and diverse videos.

The core of Seedance1.0 pro prompt writing is to "intuitively express the effect of what you want". You can use natural language to freely describe actions,/camera movements/styles/perspectives, to effectively control the video effects.

If the model does not respond well to the prompt, try to make the prompt more straightforward and easy to understand.

Examples

1. Action Prompt

Fundamental actions

  • The kitten yawns at the camera.

  • A woman is walking on the streets of Shanghai at night.

Multi-character, multi-action

  • At a rock-band live performance, the lead singer sings on stage while the guitarist, bassist, drummer, and keyboardist play energetically.

  • In an office pantry, colleagues chat during a break: one animatedly recounts a weekend story, another doubles over laughing, a third presses for details, while others chime in.


2. Lens Language

Basic camera movements

  • At a fashion show, the camera pans right to film a gorgeously dressed model walking the runway in profile.

  • In a courtroom, a zoom lens pushes in on a defendant’s nervous face while the background pulls away, heightening tension.

Complex camera movements

  • A girl plays with her puppy on a living-room carpet. Starting at ground level, the camera follows the puppy, tilts up to the girl’s smile, performs a close 360-degree pan, then slowly zooms in as she lifts the puppy.

  • A woman stands by a bright window holding coffee. The camera slides in over her shoulder to capture steam, glides through the window for a brief street view, rotates 180° back inside, then pulls away, all in one continuous shot.


3. Shot Size & Perspective Control

  • A train exits a tunnel; exterior view shows rough walls flashing past, then a burst of brightness on exit.

  • High-altitude aerial view: a camel caravan crosses a vast desert, emphasizing the journey’s hardship.

  • Macro: a caterpillar crawls along a petal, its body hairs clearly visible.

  • Over-the-shoulder: in a living room, a father teaches his son chess, capturing the son’s thoughtful face and board layout.

  • “Through-the-box” shot: two people peer inside; one reaches in and lifts out a kitten.


4. Multi-stylized Videos

  • Black-and-white line-drawing style: a girl walks right through a sketch forest.

  • A felt kitten walks down a clay street.

  • 3-D cartoon: a personified horse attends class in a schoolroom.

  • Japanese-manga style: a stylish woman wearing sunglasses takes a selfie on a Tokyo street.

  • Voxel style: a robot sits astride a rocket.

  • American-comic style: a muscular man lifts weights.


5. Prompt Controls — Character Appearance

  • A beautiful Chinese woman in an elegant black cheongsam smokes in a Western-style living room.

  • A slightly chubby young woman stares at the camera; she has three-white eyes, a mole near one eye, rough skin, and red-and-blue lighting on her face.

  • A disheveled man hungrily eats a chicken drumstick in a bare room.

  • A girl takes a selfie, flashing a “scissors” hand sign.

  • A 300-pound white man curls up on a sofa watching TV, screen light rippling across his face.


6. Prompt Controls — Aesthetic Feel of the View

  • Two hands clasp; the scene feels like a festive, rustic short video.

  • Two hands intertwine; framed as a European art-house film.

  • Two hands clasp; styled as a retro Hong-Kong film.

  • Two hands intertwine; presented as a horror movie.

  • Oil-painting film style: in the English countryside, a blond woman in a knit sweater meets a handsome man’s soulful gaze.

  • Retro textured street scene: a violinist plays under neon lights at night.

  • 1980s TV-drama look: a man writes beneath a table lamp, costumes and makeup intentionally dated.

  • Micro-world fantasy: on garden leaves, pixies appear; camera pans right as the hero in a petal cloak steps out holding a glowing-gem wand.


7. Multi-lens Capability

  • American-comic 2-D animation: close-up of a handsome young man stretching and yawning → cut to a woman filming him as he props elbows on knees → cut to a top-down shot of a steaming coffee cup set on a magazine.

  • Close-up of a red-haired girl’s astonished face → cut to a flowerpot on a ruined windowsill → overhead shot as she walks to it → close-up of her eyes, then her lips as she whispers the plant’s name.

  • Dilapidated factory crime scene: low-angle shot of a detective walking → pan to level shot of cluttered machines → zoom and slight tilt onto a bloody footprint, building suspense.

  • Futuristic lab: panoramic shot of a quantum computer and scientist on a holographic screen → close-up of the computer as red light erupts → low-angle close-up of the scientist’s face bathed in red.


8. Creative Special Effects

  • Night at New York Harbor: the Statue of Liberty ignites at the base and rockets skyward, flames lighting the city and sea as the camera tracks her fiery ascent.

  • A bullfrog lounges on a pink massage chair while a white cat kneads its shoulders like a pro masseur.

  • A tomb robber’s skeleton drags itself across blue bricks in a candle-lit tomb, shards of porcelain and coins scattered around, the skull grinning as it crawls.

Seedance 1.0 lite prompt guide

The Seedance 1.0 Lite model, as a small parameter version of the model series, not only achieves excellent video generation quality, but also significantly improves the generation speed, making it a cost-effective choice that balances effectiveness and efficiency.

1 · Camera-movement prompts

Camera switching

  • Kittens and puppies eat cat food, and the camera switches to a close-up of the cat food.

  • The watch pointer rotates at a constant speed; the camera switches as a man adjusts his gold-rimmed glasses, soft light glinting on the dial.

  • Panoramic shot: a smiling model walks toward the camera; camera switches to a close-up of the model’s lower body, highlighting the straight-cut pants and fabric drape while she walks.

Move the camera (up / down / left / right)

  • The camera moves slowly down toward the street, revealing glowing streetlights.

  • Deep inside a temple, a backpacker discovers a statue of an ancient sage; the camera moves left to reveal the statue holding an ancient book, hinting at guarded knowledge.

Zoom & close-up

  • A family is camping in a cheerful atmosphere; close-up shots of the tent, followed by a 360-degree panorama of the campsite.

  • Close-up of a sculptor’s trembling hands chiseling marble; the camera then zooms out to reveal a child sculptor wearing a confident smile.

Surround / 360-degree

  • The camera wraps around the subject and slowly pushes in to a close-up of the character’s face.

Aerial

  • A man spreads his arms wide as an aerial shot circles around him.

Hand-held

  • A hand-held camera jitters slightly while following a cat strolling through a rose garden.

Follow

  • The camera follows a man riding a Harley motorcycle in close-up; he grins wildly, the camera jolts upward, and a vulture is revealed circling overhead.


2 · Degree-adverb prompts

  • A vehicle quickly moves forward through a time-space tunnel; light and shadow streak past as the camera rapidly follows.

  • An athlete in professional sportswear sprints: legs move quickly, arms swing powerfully; after he crosses the finish line, the crowd erupts in cheers as a follow-shot captures every detail.


3 · Detailed t2v examples

  • Psychedelic portrait – Cool light-blue tones, butterfly lighting; close-up of a young white woman with short black hair, arched right eyebrow, thick lashes, high nose bridge, and biting red lips. She stares at the camera disdainfully; the camera pulls back as floating shards of broken glass partially obscure her face.

  • Winter-valley close-up – In a snow-filled valley, a beautiful young woman in a black robe turns to look at the camera. Snowflakes drift past lush green vegetation behind her; the camera slowly circles left as she raises her manicured hand to her chin and smiles seductively.

Hailou 01-Director prompt guide

1 · Camera-movement prompt examples

(ready-to-copy text-to-video prompts)

  1. [Push in] A lamb stands alone in a snowy field, snowflakes gently falling around it.

  2. [Pan left, Pan right] A woman in business attire sits at a café, sipping coffee while working on her laptop.

  3. A man picks up a leather-bound book from a dusty shelf [Pedestal up], then begins reading it by candle-light [Static shot].

  4. Above an ancient battlefield [Pan right, Pedestal down], a close-up reveals a warrior woman’s face, dirt-streaked but determined.

  5. [Tracking shot] A red fox moves stealthily through autumn woods, hunting for prey.

  6. [Push in, Zoom out] A surprised woman stands in a crowded beach scene, her expression frozen as the background comes into focus.

  7. The shot begins from a dark rooftop [Pedestal up, Pan left], revealing a glittering cityscape with a giant red moon hanging in the night sky.

  8. [Pan left, Tilt up] Starting from a small rabbit looking upward, following its gaze to a bright red balloon drifting into the sky.

  9. [Pull out, Tilt up] A knight stands amid ruins with a broken sword; the camera reveals burning city walls and a vast dragon shadow overhead.

  10. [Pan left, Pan right, Tracking shot] A woman in a flowing red dress strides confidently through a modern subway station. Replicate


2 · Supported camera-movement tags

Category

Bracket tag(s)

What it does

Horizontal

[Truck left] / [Truck right]

Entire camera slides sideways

[Pan left] / [Pan right]

Camera pivots left/right on its axis

Vertical

[Pedestal up] / [Pedestal down]

Camera rises or lowers vertically

[Tilt up] / [Tilt down]

Camera tilts up/down while staying put

Depth

[Push in] / [Pull out]

Camera moves forward/backward

[Zoom in] / [Zoom out]

Focal length change (lens zoom)

Special

[Tracking shot]

Follows the subject’s motion

[Shake]

Adds handheld-style vibration

[Static shot]

Locks the camera in place


3 · Preset shot-combination shortcuts

Name

Bracket combo

Typical use

Left circling

[truck left, pan right, tracking shot]

Orbit counter-clockwise around a subject

Right circling

[truck right, pan left, tracking shot]

Orbit clockwise

Left walking

[truck left, tracking shot]

Lateral follow shot (subject walks)

Right walking

[truck right, tracking shot]

Opposite side

Upward tilt

[push in, pedestal up]

Glide in while rising

Scenic reveal

[truck left, pedestal up]

Lateral slide plus craning up

Stage shot (R/L)

[pan right, zoom in] / [pan left, zoom in]

Dramatic spotlight pan + zoom

Downward tilt

[pedestal down, tilt up]

Descend while tilting to keep focus


Quick best-practice reminders

  • Max three movements per bracket set for clean results.

  • Place each bracket exactly where the movement starts in your sentence.

  • Mix natural language with bracket tags for precise yet readable prompts.

  • Combine sequentially by adding a new bracketed set later in the same prompt.

Did this answer your question?