Muse Video Prompts: 7 Examples and a Simple Formula

Learn to direct motion, camera movement, and native audio in Muse Video with seven short video prompts and a simple writing formula.

Generate a video Explore examples ↓

Muse Video Prompt Guide title over a sunlit underwater reef — A reef scene gives the camera a foreground to pass and a school of fish to follow.

A still frame establishes the scene. A video prompt gives the scene a direction, from the first visible movement to the last sound the viewer hears.

How to write a Muse Video prompt

Start with one visible change. Decide what the viewer sees at the beginning, what moves, and where the shot ends. Then describe the sound that belongs to the action.

Subject: Name the person or object and one detail that should stay recognizable.
Action: Describe what changes and where the movement ends.
Place and light: Set the scene and say where the light falls.
Camera: Give the starting view and one camera move, or ask for a locked shot.
Pacing: Give the action time to start, finish and settle.
Continuity: Name the parts of the subject or setting that must stay consistent.
Sound: Name the source of each important sound and when it should happen.

Worked example · A kettle on the stove

A stainless steel kettle sits on a gas stove in a sunlit kitchen. Hold a side close-up as a thin thread of steam thickens above the spout. Keep the handle and stove controls in frame. Hear the burner hiss first, followed by a brief whistle.

The kettle anchors the frame, the steam supplies the visible change, and the whistle gives the shot a clear ending.

Choose an input

Text to video

Establish the subject, setting and starting frame in words. Describe the motion and audio once the scene is clear.

Image to video

Use the uploaded image to establish appearance. Tell Muse Video what to animate and what to preserve, such as a watch face or the direction of light.

For a reference-led shot, identify what the image controls and describe any motion in the text. For spoken lines, name the speaker and place the line after a visible cue.

Read the Muse Video User Guide →

Video prompt examples

Each example is planned as a 10-second clip. Set the input mode and aspect ratio shown above the prompt. Use the time ranges as pacing cues when reviewing the result.

Backstage before the curtain

Text to video · 16:9

Prompt 01

Create a 10-second photorealistic 16:9 clip in one continuous shot. A violinist in a black concert jacket waits backstage beside a red velvet curtain.

0-3s: Frame the violinist from the side, holding the violin low. A narrow strip of stage light crosses their face.
3-7s: Dolly slowly alongside as they take one breath, lift the violin to their shoulder and step into the light.
7-10s: Stop at the curtain edge and hold on their profile as the bow settles onto the strings.

Visual style: Warm stage light against a cool, dim backstage corridor. Natural skin texture and deep red velvet.
Native audio: A muffled audience murmur, soft shoes on wood, and one clean violin note when the bow moves.
Constraints: Keep the same instrument, jacket and performer. No cuts or music bed. No text overlays or background music.

Try Muse Video →

What to check · Check that the first note begins with visible bow movement.

Tailor fitting a jacket

Text to video · 16:9

Prompt 02

Create a 10-second photorealistic 16:9 clip in one continuous shot. A tailor checks the fit of an olive jacket on an adult customer in a small tailoring studio.

0-3s: Open at chest height with both people visible. The customer stands facing a mirror while the tailor studies the jacket shoulder.
3-7s: Slide gently to the side as the tailor smooths the shoulder seam with one hand. Keep the jacket and its reflection visible.
7-10s: Hold as the customer turns slightly and the tailor steps back to inspect the fit.

Visual style: Soft window light, muted olive fabric and realistic mirror reflections.
Native audio: Quiet room tone, fabric brushing under the hand and a soft footstep as the tailor steps back.
Constraints: Keep the jacket seams and buttons consistent. The reflection follows the customer exactly. No text overlays or background music.

Try Muse Video →

What to check · Check that the fabric moves with the tailor's hand and the reflection matches the turn.

A bookstall in the rain

Text to video · 16:9

Prompt 03

Create a 10-second photorealistic 16:9 clip in one continuous shot. A traveler in a tan coat stops at a covered outdoor bookstall on a rainy afternoon.

0-3s: Follow just behind the traveler as they step under the striped awning. Keep their profile and the wet pavement visible.
3-7s: Move alongside as they pick up one blue hardcover and open it under the awning light.
7-10s: Hold at a three-quarter angle as they turn one page and glance out toward the rain.

Visual style: Diffused gray daylight, warm stall lighting and natural reflections on the pavement.
Native audio: Rain pattering on the awning, a faint street wash and one crisp page turn.
Constraints: Keep the same book, coat and stall layout. No legible book text. No text overlays or background music.

Try Muse Video →

What to check · Check that the book stays recognizable and the page-turn sound lands on the gesture.

Skateboard quick hook

Text to video · 9:16

Prompt 04

Create a 10-second photorealistic 9:16 clip in one continuous shot. An adult skateboarder in a burgundy sweatshirt rolls down an empty concrete path.

0-3s: Open low beside the board in a vertical frame. Show the wheels already moving and keep the rider visible above them.
3-7s: Track alongside at the same speed as the rider crouches slightly and rolls over a painted line. Rise toward waist height.
7-10s: Hold as the rider coasts upright into an open patch of sunlight.

Visual style: Bright late-afternoon light, gritty concrete texture and restrained motion blur.
Native audio: Steady wheel rumble and a short click as the board crosses a pavement seam.
Constraints: Keep both feet on the same board after the opening. No tricks, cuts or added people. No text overlays or background music.

Try Muse Video →

What to check · Check that the rider stays inside the vertical frame and the wheels maintain contact.

A field recorder demonstration

Text to video · 16:9

Prompt 05

Create a 10-second photorealistic 16:9 clip in one continuous shot. A sound recordist demonstrates a compact unbranded recorder beside a quiet stream.

0-3s: Open on the recordist crouching near the bank with the recorder visible in both hands.
3-7s: Push closer as they raise the recorder toward the water and press its side button once. Keep the button and fingertip in focus.
7-10s: Hold the recorder in the foreground with the recordist visible behind it and moving water softly blurred.

Visual style: Natural green reflections, soft outdoor light and a crisp matte finish on the recorder.
Native audio: Steady water and a small tactile click at the button press.
Constraints: Preserve the recorder shape and button layout. No logos or screen overlays. No text overlays or background music.

Try Muse Video →

What to check · Check that the button press is readable and that its click is distinct from the stream.

One line at the greenhouse

Text to video · 16:9

Prompt 06

Create a 10-second photorealistic 16:9 clip in one continuous shot. A greenhouse keeper in a slate-blue apron notices the first flower opening on a potted plant.

0-3s: Start at eye level as the keeper leans over a bench and gently parts two leaves.
3-7s: Push in slightly as they spot the bloom and smile. The keeper looks toward someone off camera and says, “It opened this morning.”
7-10s: Hold on their expression and the bloom after the line, with the leaves settling back into place.

Visual style: Soft light through greenhouse glass, realistic foliage and natural facial movement.
Native audio: A leaf rustle before the single spoken line, with light greenhouse ambience underneath.
Constraints: Keep one visible speaker and one short line. No subtitles or additional voices. No text overlays or background music.

Try Muse Video →

What to check · Check that the speaker's mouth matches the line and the leaf rustle precedes it.

Character continuity

Image to video · 16:9 · Upload a portrait with the subject seated by a window

Prompt 07

Portrait used for the character continuity example — Use the same portrait for generation and the guide.

Create a 10-second photorealistic 16:9 portrait clip using the uploaded image as the opening frame.

0-3s: Match the portrait composition. Hold on the seated subject in soft window light while a curtain moves gently in the background.
3-7s: The subject slowly shifts their gaze from the window to a book already resting in their lap. Ease the camera slightly closer at eye level.
7-10s: Hold on the subject as they settle into reading. Keep the book and a strip of window light in the frame.

Visual style: Natural skin texture, soft daylight and gentle background blur.
Native audio: A faint curtain rustle and quiet room ambience.
Constraints: Preserve the face, hairstyle and outfit from the uploaded portrait. No cuts, dialogue, text overlays or added people.

Try Muse Video →

What to check · Compare the face, hairstyle and outfit with the portrait throughout the gaze shift.

Text-only prompt exercises

Keep the viewpoint attached to the subject

Starting draft · A camera glides through a greenhouse.

Attach the camera to a slow-moving garden cart and keep the edge of its wooden tray in the lower right. Glide along a greenhouse aisle past rows of ferns and stop beneath the skylight. Hear the wheels rolling and water dripping from the leaves.

The tray should move with the viewpoint, and the cart should stop beneath the skylight.

Connect the beats of a short story

Starting draft · A storm rolls in.

Lock the camera on a shallow puddle reflecting an old brick wall. A few raindrops break the reflection, then the rain picks up and overlapping ripples fill the frame. Hold on the disturbed water. Hear isolated taps build into a steady patter.

Each beat changes the same reflection, and the sound builds with the rain.

Give the image and the prompt separate jobs

Starting draft · Make this photo cinematic.

Use the uploaded photo to establish the reading chair, window and afternoon light. Keep the room and camera fixed. Animate only the curtain as it lifts in a brief breeze, then settles against the sill. Hear the fabric rustle and a distant passing car.

Compare the first and last frames with the uploaded photo. The chair and framing should stay in place.

Prompt direction quick sheet

Direction	How to use it
Locked shot	Keep the frame fixed while the subject moves.
Slow push-in	Move toward the subject and stop when the important detail fills the frame.
Side tracking	Move alongside the subject at its speed and keep it in the same part of the frame.
Reveal	Start with part of the scene hidden and name what the camera uncovers.
Reaction	Name the visible cue, then describe the look or gesture it causes.
Audio cue	Link the sound to the action, such as a click when the latch closes.
Final hold	Describe what remains visible after the movement stops.

Common mistakes and fixes

If the shot feels rushed, keep one main action and give it a clear ending.
If the camera moves in the wrong direction, state what stays in frame and where the camera ends.
If the sound feels disconnected, tie it to a visible event, such as the bell ringing as the door opens.

Bring the shot together

Picture the opening and closing frames. Check that the action can take the subject from one to the other.
Read the camera and sound directions together. Check that each important cue is visible or has a clear source.
Watch the whole generated clip with sound, then revise the one instruction that would make the biggest difference.

Try a prompt in Muse Video

Frequently asked questions

How detailed should a Muse Video prompt be?

Give enough detail to picture the shot. Describe the subject, visible action, setting, camera and sound. Add a continuity instruction when something specific should stay the same, such as a jacket color or a reference portrait.

Does every video prompt need timestamps?

A single continuous action can usually be described in one paragraph. Use time ranges when several actions or audio cues need to fit into one clip. Treat those ranges as pacing guidance and check the actual timing after generation.

How can I shorten a long prompt?

Keep the main action, camera direction, ending and important sound cues. Cut repeated mood words and background details that do not affect what the viewer sees or hears.

What does an AI video prompt control?

It describes what to show, how it should move and how the camera should follow it. In Muse Video, you can also describe what the viewer should hear, such as footsteps, a doorbell or a short line of dialogue.

How do I start writing a video prompt?

Start with one visible action, such as a barista pouring espresso over ice. Add the setting and framing, then describe where the shot ends and when the key sound happens.

Can I use a prompt with a reference image?

Yes. Let the image establish the appearance and opening composition. Use the prompt to describe what changes and which details should remain consistent. Example 07 shows how to direct a gaze shift while preserving the face, hairstyle and outfit.

Can a prompt-writing tool help me draft a video prompt?

Use it to turn a rough idea into a first draft or find missing details. Before generating, check that the camera can see the action, the timing fits the clip, and the text agrees with any reference image.