Seedance Prompt Examples, by Use Case

Placeholders are in {curly braces}. In the AiuniVid workbench, choosing a saved character inserts its tag for you, numbered in the order you pick; on the Open API, write @image1, @video1 and so on in the order the files are attached.

1. A vertical story shot with a recurring character

Settings: reference to video, 9:16, five to fifteen seconds per shot. Use Seedance 2.5 when a take runs past 15 seconds or needs more than 9 references.

Vertical 9:16, {10}-second shot, {stylized 3D animation}.
@image1 is the main character only. Keep face, hair and outfit identical.
@image2 is the product only. Keep its shape, colour and logo unchanged.
Scene: {a small apartment hallway, evening light}.
Action: {the courier} {hands over the parcel}, then {steps back, surprised}.
Camera: slow push-in on {the courier's face}, then a quick whip-pan to {the parcel}.
Dialogue: {NEIGHBOUR}: "{You're not the usual driver.}" Only {NEIGHBOUR} speaks.
No subtitles, no on-screen text.

• Each reference has one job, stated in the prompt. An image that is only the character and an image that is only the product are far less likely to bleed into each other.
• One beat, one camera move. The next beat is the next clip; the story is assembled in the edit.
• Name the speaker and quote the line. Seedance generates the audio with the picture, and this helps the voice land on the right face.
• Watch out: a photorealistic character, 3D renders included, can be refused as a real face. Save the character under Assets first and reference it from there. The whole workflow is in keeping a character consistent across shots.

2. An opening hook

Settings: text to video, 9:16, four to five seconds. Draft on Seedance 2.0 Fast or Mini, then render the keeper on 2.5 or 2.0.

Vertical 9:16, {5}-second shot, {cinematic live-action look}.
Start in the middle of the moment, no establishing shot.
{A young chef} {drops a plate on the kitchen floor} during {opening night}.
{The head chef} {turns around slowly}.
Camera: handheld, fast push-in on {the young chef's face} in the last second.
{HEAD CHEF}: "{Again.}"

• Start inside the conflict. The first second decides whether a short gets watched, so the prompt opens mid-moment rather than on a wide establishing shot.
• One action and one reaction. Without a reference image, describe each person with two or three concrete traits; once a character has to recur, move them into reference mode.

3. A motion-reference demo: drills, exercises, dance

Settings: reference to video, 16:9 or 9:16, five seconds, one character plus one reference video of at least 2 seconds.

{16:9}, {5}-second single continuous take, real-time speed, no cuts.
@image1 is the only person in the video. Keep identity, outfit and proportions.
@video1 is a motion reference only. Copy the movement, timing and footwork,
not the person or the background.
Action: exactly one {kettlebell swing}, from {a hinged start position} to {a standing lockout}.
Camera: locked, eye level, {three-quarter front} view, full body in frame.
Setting: {plain studio floor}, even light.

• Say which attachment does what. The video supplies motion only; the image supplies identity. BytePlus documents this pairing as motion plus subject reference.
• Exactly one movement. A combination packed into a five-second take tends to blur the moves together. Render combinations as separate clips.
• Left and right get mirrored. In a front view the subject's right hand appears on the left of the frame. Write "the subject's right hand" rather than "the right hand".

4. Storyboard to a cinematic shot

Settings: reference to video, 16:9 or 21:9, eight to ten seconds. Seedance 2.5, because storyboard, characters and environment together use several reference slots.

{16:9}, {10}-second cinematic shot, one continuous take.
@image1 is the storyboard. Use it for plot, composition and scene order only.
@image2 is {the knight}: keep face, armour and proportions.
@image3 is {the frozen bridge}: keep layout, light and palette.
Scene: {the knight crosses the bridge at dawn as the ice starts to crack}.
Camera: {slow tracking shot from behind, then a low-angle reveal of the far gate}.
Style: {stylized 3D game cinematic}, {cold blue light, drifting snow}.
Generate one continuous scene, not a collage of panels.

• The storyboard is read loosely. BytePlus describes a multi-panel storyboard as a high-level plot reference, recommends 15 panels or fewer in simple line art with no text on the panels, and asks for actions and camera movement in the prompt.
• Attach characters and the environment separately and label the storyboard as plot only.
• Need exact start and end frames? Use first and last frame in image to video instead of a storyboard.

5. A product clip from one still

Settings: image to video, the same aspect ratio as the photo, about five seconds.

Use the uploaded image as the first frame.
Keep the {ceramic mug} unchanged: shape, colour, material, logo and printed text.
Motion: {a slow half turn on the counter while morning light moves across the glaze}.
Camera: {static, slight push-in}.
Background: {keep the kitchen counter; add soft steam rising from the mug}.
No new text, no extra products, no hands.

• Lock what must not change. The model treats everything in the frame as something it may animate, and printed text drifts first.
• Describe only the motion around the product — light, a turn, steam — rather than the product itself, which the image already defines.
• Need a predictable ending? Upload a second image as the last frame.

Settings at a glance

For how the two main models differ on length, references and cost, see Seedance 2.5 vs 2.0. Every quote appears before anything is charged; see credit pricing.

Trying a template on Seedance 2.5 for the first time? Run it as a 480P draft: it shows whether the prompt lands before you pay for 1080P, and the final is rendered from the draft you liked. See Seedance 2.5 draft mode.

Frequently asked questions

Sources

• BytePlus ModelArk · Dreamina Seedance 2.5 prompt guide
• BytePlus ModelArk · Dreamina Seedance 2.5 tutorial