Directory Image
This website uses cookies to improve user experience. By using our website you consent to all cookies in accordance with our Privacy Policy.

Seedance 2.5 Prompting Playbook

Author: Aivon Ai
by Aivon Ai
Posted: Oct 17, 2026

Chapter 1 — Quick Start: What Seedance Can Do

Seedance is ByteDance Seed team's multi-modal video generation model. Seedance 2.0 and 2.5 are fully integrated into viddo.ai. The capability matrix centers on four pillars: long-form narrative, strong reference, precise editing, and multilingual fluency. If you only remember one line about it:

Seedance turns natural language + multi-modal references into 4–30 second videos with full control over shot, sound, subtitles and storyboard, plus post-production precision editing.

1.1 Capability Snapshot

Capability

Seedance 2.0

Seedance 2.5

Single-clip duration

Up to 15s

4–30s at fixed 24fps

Multi-modal references

Supported

Up to 50: 30 images + 10 videos + 10 audio

First/last-frame generation

Supported (parameter-controlled)

Supported + multi-keyframe sequence control

Video editing (add/remove/modify)

Supported

Supported, higher precision

Forward / backward extension

Supported

Supported, tighter aspect-ratio & duration locking

One-click reel / seamless transition

Basic

Full

Gray-model reference / render

Not supported

Coarse + fine-grained gray models

Multilingual prompts

Supported

Native 10+ languages

1.2 Typical Use Cases

Seedance is not a toy. In production it shines for:

  • Brand TVC commercials (15–30s, single-take)
  • Short drama / cinematic shorts (one-shot, multi-character micro-expressions)
  • E-commerce SKU showcase videos (multi-angle, multi-shot)
  • Education / explainers (step-by-step, biological growth, cultural heritage)
  • 3D / gray-model rendering (architecture, product)
  • Film pre-visualization (camera pre-viz, age progression, cross-style variants)
1.3 Reader Mindset

Think of Seedance 2.5 as a visual content producer — you (the director) supply the script, it shoots it. Everything that follows is essentially teaching you to write the script so the model shoots it right.

Chapter 2 — Text Prompt Basics2.1 Universal Formula: Six Building Blocks

Seedance understands natural language directly. There is no rigid template, but a battle-tested formula helps you organize your thoughts:

Subject + Action or Event + Scene & Environment (optional) + Visual Style (optional) + Camera & Editing (optional) + Sound (optional)

Element

Purpose

Writing Tips

Subject + Action

Defines who is doing what

Describe the main motion in broad strokes; only spell out key actions in detail. Avoid repeating the same action.

Scene & Environment

Location, time, spatial relationships, background

Use visual, concrete keywords: dawn / rainy night / glass greenhouse / wooden workbench

Visual Style

Lighting / color / material / texture / mood

Documentary realism, cinematic grading, shallow DOF, soft light

Camera / Editing

Shot size, position, motion, focus, cuts

Push / pull / pan / track / dolly / orbit / dive / pull-back / tilt / handheld

Sound

Dialogue, voice timbre, ambience, SFX, music

Use special markers (see §2.4)

2.2 Base Template

performs in.

Visuals:.

Camera:.

Sound:.

2.3 Worked Example: Potter Making a Cup

A potter in a sunlit morning studio finishes a pale-blue ceramic cup, lifting it off the wheel and placing it at the center of a wooden shelf.

Soft morning light streams through the window; the damp clay shows delicate sheen; the workbench stays tidy.

Camera starts with a medium shot of the wheel-throwing motion, then slowly pushes in to the cup surface texture, and finally cuts to a head-on shot of the shelf.

Keep the wheel's low hum, the clay-rubbing sound and a quiet indoor ambience.

2.4 Special Markers for Sound and Text

Prompts can use plain language. To explicitly separate music, SFX, dialogue and subtitles, use these markers:

Content

Marker

Example

Music

`()`

`(gentle piano music in the background)`

Sound effects

``

``

Dialogue

`{}`

`{hello, welcome back}`

Subtitles

`【】`

`【Chapter 1: Departure】`

2.5 Negative Controls (What NOT to Generate)

No background music. Keep only dialogue, ambience and action SFX.

No subtitles.

No sound at all.

2.6 Language Reinforcement for Dialogue

When the dialogue text is not the model's default (Chinese), or you need to force English/Japanese/Korean, strongly recommend specifying the language in front of the line. Recommended formula:

Dialogue language + regional variant or accent + delivery style + speaker + {line}

Dialogue language: American English. The girl says naturally and casually in American English: {I thought you weren't coming.}

Dialogue language: authentic Los Angeles American English. The young man speaks with natural LA slang: {No way, you actually made it.}

The girl whispers softly in Japanese: {もう大丈夫です}

Chapter 3 — Multi-Modal References: Pinpointing @Image / @Video / @Audio3.1 Why "Precise Referencing" Is the Core Rule

Seedance accepts any combination of text + images + videos + audio. The more assets you supply, the more critical it becomes that the model knows exactly what each asset is for — a hard rule repeated across S1, S2 and S3:

Mapping between assets and on-screen elements must be written into the prompt. Do not rely on text labels baked into images, and never let the model guess which asset corresponds to which character, prop or scene.

3.2 Input Quantity and Recommended Ranges

Seedance 2.5 accepts up to 50 assets per task. Going beyond the recommended range is possible but stability drops:

Asset Type

Hard Limit

Recommended Range

Stretch Range

Images

Up to 30, each ≤ 4K

Subject images: 1–8 subjects

9–12 subjects

Videos

Up to 10, total runtime ≤ 30s

Subject A/V: 1–5 subjects, each 5–10s

6–10 subjects

Audio

Up to 10, total runtime ≤ 30s

Only task-relevant dialogue, voice, ambience or music

  • p>
Video editing

Video + reference images

Original video ≤ 20s, references 1–5 images

6–8 images

3.3 General Sentence Pattern

Reference / Extract / Combine / Follow + Image n + the referenced element, generate + the scene description, keep + the referenced element + consistent

Common verbs: reference / extract / combine / follow / lock / preserve / exclude

3.4 Documenting Each Asset's Role

For every asset, do two things: (1) describe what attribute it provides, (2) when the asset risks being "carried in" wrongly, write down what to exclude. S3's recommended template:

@Image1 is used for 's.

@Video1 is used for.

@Audio1 is used for 's.

performs in.

Visuals:. Camera:.

3.5 Worked Example: Potter

@Image1 is used for the potter's facial features, hair and deep-green apron. Do not use its background.

@Image2 is used for the studio's wooden workbench, window placement and morning light. Do not use any person in this image.

@Video1 is used for the hands-on-wheel motion, lifting the cup and placing it. Do not use the person, clothing or scene from the video.

The potter finishes a pale-blue ceramic cup in the morning studio, lifting it off the wheel and placing it at the center of the wooden shelf.

Camera starts with a medium shot of the wheel-throwing motion, then slowly pushes in to the cup surface texture. Keep the wheel hum, the clay-rubbing sound and quiet indoor ambience.

3.6 Multi-View: Different Angles of the Same Subject

When multiple images show different angles of the same subject (person or product), explicitly label them as "the same X from different views" so the model doesn't treat them as distinct subjects.

@Image1 defines the front view of the same folding lamp.

@Image2 defines the left side structure of the same folding lamp.

@Image3 defines the right side structure of the same folding lamp.

@Image4 defines the rear structure of the same folding lamp.

All four images together define a single folding lamp; the output must contain only one folding lamp.

Chapter 4 — On-Screen Text: Slogans, Subtitles and Speech Bubbles

Seedance 2.0 onwards supports on-screen text generation across T2V, I2V, R2V and V2V scenarios. The model auto-matches style and color, and also accepts explicit specifications for color, style, timing, placement and behavior.

Prefer common characters. Avoid rare characters and special symbols for best rendering.

4.1 Slogans (Opening or Closing Tag)

Formula: "text content" + "appearance timing" + "position" + "appearance behavior", "text traits (color, style)"

Hand-drawn comic style. Three people sit together eating fried chicken from Image 1 in a warm friendly atmosphere. The frame gradually blurs and the text "Happiness Lives in Seedance" appears in the center.

  • Seedance 2.0 提示词指南 _ External Seedance 2.0 Prompt Guide\图片和附件\02177105639618900000000000000000000ffffac177d1a635886.mp4

Slogan example - "Happiness Lives in Seedance"

  • p>4.2 Subtitles (Voiceover or Dialogue)
Subtitles appear at the bottom of the frame. The text content is "...". Subtitles must stay perfectly in sync with the audio rhythm.

Example 1: Voiceover

Generate a video with voiceover: a deep, calm male voice says: "In the grand cosmos, our world is only a brief moment. Yet within it, life thrives against all odds."

The scene slowly transitions from night to dawn. Stars fade away, the sun rises from behind the mountains. Subtitles appear at the bottom of the frame in sync with the dialogue.

  • Seedance 2.0 提示词指南 _ External Seedance 2.0 Prompt Guide\图片和附件\02177097189398600000000000000000000ffffac14520e3267e1.mp4

Voiceover subtitle example

  • p>
Example 2: Dialogue

The two people from Image 1 chat in an office. The woman speaks first: "You always arrive just on time. Do you enjoy that perfect timing?"

The man smiles and replies: "I have my own rhythm." Their conversation flows naturally; matching subtitles appear at the bottom of the frame.

  • Seedance 2.0 提示词指南 _ External Seedance 2.0 Prompt Guide\图片和附件\02177112867931200000000000000000000ffffac14bbb053141a.mp4

Dialogue subtitle example

  • p>4.3 Speech Bubbles
says: "...". As they speak, a speech bubble appears around them containing the dialogue.

Example 1: Running Track Conversation

The two people from Image 1, wearing sportswear, run on the school track. The girl looks at the boy and smiles confidently: "We can definitely do it!"

Cut to a close-up of the boy; he hesitates: "Are you sure?"

Cut back to the girl's medium close-up; she answers brightly: "Yes!" The mood is bright and decisive. Speech bubbles with the matching dialogue pop up around each speaker.

  • Seedance 2.0 提示词指南 _ External Seedance 2.0 Prompt Guide\图片和附件\02177122447710600000000000000000000ffffac177d1a42e779.mp4

Speech bubble example - running track

  • p>
Example 2: Strawberry Picking

Using the girl from Image 1 and Image 2 as reference, the girl picks a strawberry in a strawberry garden, takes a bite and smiles: "This is the real deal!"

A speech bubble with the line appears around her.

  • Seedance 2.0 提示词指南 _ External Seedance 2.0 Prompt Guide\图片和附件\02177122475078800000000000000000000ffffac15bbe65a3e2f.mp4

Speech bubble example - strawberry garden

  • p>Chapter 5 — Keyframes, First/Last Frames, Gray Models and Storyboards5.1 First/Last Frame Generation
In multi-modal reference mode, you can simply write `@Image1` as the first frame and `@Image2` as the last frame at the top of the prompt — no need to switch to a dedicated first/last-frame mode. The system locks the output aspect ratio to the first frame's ratio and lets you set the duration in the UI/API. First and last frames must share the same aspect ratio; otherwise the last frame may stretch.

Document each anchor image separately — do not collapse them into "Image 1 and Image 2 as first/last frames". Other reference images only supply the specified attributes; they do not replace the first/last frame composition.

@Image1 as the first frame, defining the starting composition, subject position, pose, prop state, scene and camera direction.

@Image2 as the last frame, defining the ending composition, subject position, pose, prop state, scene and camera direction.

@Image3 is used for 's. Do not alter the first-frame composition from @Image1, nor the last-frame composition from @Image2.

@Image4 is used for 's. Do not alter the first-frame composition from @Image1, nor the last-frame composition from @Image2.

.

The scene starts naturally from @Image1's first frame, flows through the continuous motion and arrives at @Image2's last frame.

Keep consistent between first and last.

Example: Perfume Atelier

@Image1 as the first frame, defining the starting composition, character position, pose, tabletop prop state and camera direction of the perfume atelier.

@Image2 as the last frame, defining the ending composition, character position, pose, tabletop prop state and camera direction of the perfume atelier.

@Image3 is used for the perfumer's face, hair and deep-green apron. Do not alter the first-frame composition from @Image1, nor the last-frame composition from @Image2.

@Image4 is used for the glass perfume bottle's shape, material and label placement. Do not alter the first-frame composition from @Image1, nor the last-frame composition from @Image2.

The perfumer starts from the first-frame pose, picks up a dropper and the glass perfume bottle, drips amber essence into the bottle, gently shakes it and caps it, then places the finished product at the center of the table, finally arriving at @Image2's last frame.

Keep the perfumer's identity and clothing, the bottle's count and structure, the wooden table layout, the warm side lighting and camera direction consistent throughout.

5.2 Multi-Keyframe Sequence Control

When multiple independent images define different stages of a process, open the prompt with "Using @Image1 through @ImageN in order as keyframes", then describe the state each image corresponds to. Independent keyframes usually align better than grid images, but they only control stage order and key states — not frame-by-frame replication.

Use @Image1 through @ImageN in order as keyframes.

@Image1 as the first frame, defining.

@Image2 defines the second keyframe:.

@Image3 defines the third keyframe:.

@ImageN as the last frame, defining.

The scene traverses the states defined by @Image1, @Image2, @Image3 through @ImageN in order, with natural continuous motion bridging each stage.

Keep consistent throughout.

Example: Orange Paper Plane Flight (4 keyframes)

Use @Image1 through @Image4 in order as keyframes.

@Image1 as the first frame, defining the orange paper plane resting on the left side of a classroom wooden desk, nose pointing right, in a locked medium shot.

@Image2 defines the second keyframe: the same paper plane is lifted by a hand from the desk, nose direction unchanged.

@Image3 defines the third keyframe: the same plane glides past the window, the curtain sways gently to the right.

@Image4 as the last frame: the same plane lands on the middle shelf on the right, nose still pointing right.

The scene traverses the states defined by @Image1, @Image2, @Image3 and @Image4, with continuous direction and speed.

Keep the plane's orange material, size and creases, the classroom layout, the afternoon side light and the camera axis consistent throughout.

5.3 Grid Storyboards and Multi-Grid References

Grid storyboards provide the overall story, shot order and rough composition. They are not meant for pixel-perfect replication. Keep them within 15 cells, use simple line art or clean diagrams with minimal text labels. Prompts should declare the reading order and then describe each cell's subject action, shot size or camera motion, plus final look and sound.

@Image1 provides an 's shot order and rough composition, read. Do not adopt.

@Image2 defines 's.

@Image3 defines 's.

Shot 1:.

Shot 2:.

...

Shot N:.

Final visuals use. Sound includes.

Example: Pottery Four-Grid Storyboard

@Image1 provides a four-cell pottery storyboard's shot order and rough composition, read left-to-right, top-to-bottom. Do not adopt the line-art style or text labels.

@Image2 defines the potter's face, short hair and dark-grey apron.

@Image3 defines the blue-glaze cup's body proportions, glaze color and curved handle.

Shot 1: wide shot of the quiet pottery studio; the potter sits at the wheel.

Shot 2: medium side shot of the potter's hands steadying the spinning wet clay as the cup forms.

Shot 3: close-up of fingers trimming the rim and handle joint; slip slowly slides down the fingertips.

Shot 4: medium close-up of the fired blue-glaze cup placed on the wooden shelf; the potter withdraws their hands.

Final visuals use documentary-realism texture. Keep the wheel hum, the wet clay sound and studio ambience.

5.4 Gray-Model Reference and Rendering

Gray-model references split into coarse and fine-grained. Decide whether the gray model supplies the motion skeleton or a fully built structure, then pick the matching template.

Type

Best For

Asset Requirements

Prompt Focus

Coarse gray model

Pre-visualizing action, path, blocking, camera or cuts with simple geometric placeholders

Clear geometric relationships, complete action sequence

Map each gray subject to a final subject; document which timing/space information to inherit

Fine-grained gray model

Already fully modeled; you only need to swap material, color, characters, scene or style

Complete model, clean frame

Keep structure, action and camera; document which attributes to re-render

5.4.1 Coarse Gray Model (Mobile Cart Example)

@Video1 is a coarse gray-model reference, only providing character walking path, cart movement direction, locked camera, one push-in and two cuts. Do not adopt the gray geometric look or empty scene.

The tall cylinder in @Video1 corresponds to the docent.

The cuboid in @Video1 corresponds to the mobile cart.

@Image1 defines the docent's face, blue uniform and badge.

@Image2 defines the mobile cart's white metal frame and transparent cover.

@Image3 defines the tech showroom's curved wall, grey floor and ceiling strip lights.

The docent pushes the mobile cart along the curved wall, stops at the central platform and opens the transparent cover.

Keep @Video1's walking path, subject blocking, push-in direction and cut positions.

Visuals use a bright realistic documentary style. Keep footsteps, cart wheels and the showroom ambience.

5.4.2 Fine-Grained Gray Model (Ring Installation Re-Render)

@Video1 is a fine-grained gray-model reference. Keep the ring installation's full structure, three-ring rotation, pedestal position, orbit camera and cuts. Do not adopt the original gray material or empty background.

@Image1 defines the outer ring's brushed brass material.

@Image2 defines the inner leaf layer's translucent blue glass material.

@Image3 defines the contemporary art gallery's white curved walls, deep grey floor and overhead soft light.

Re-render the ring installation in @Video1 as a dynamic sculpture of brass and blue glass; re-render the scene as a contemporary art gallery.

Keep @Video1's structure, rotation rhythm, orbit camera and cuts. Keep the installation's low rotation sound and quiet indoor ambience.

Chapter 6 — Video Editing: Add, Remove, Replace, Subject & Background Swaps

Seedance 2.0 already supports video editing; Seedance 2.5 raises precision further. Editing tasks automatically lock the output aspect ratio to the input video's and lock the duration (which you cannot override). Output duration may drift by up to ~0.3s due to input frame handling, but overall content and event order remain essentially preserved.

6.1 Universal Three-Part Template

【Edit Goal】

Edit @Video1 to across.

【Original Video Responsibility】

@Video1 is the sole editing master. It owns.

【Target Asset Responsibility】

@Image1 or @Audio1 supplies.

Throughout the extension, keep,, , and continuous.

The same subject remains a single continuous entity throughout — no duplication or splitting. The character's appearance and the object's part count remain stable.

Example: Gardener Places a Wicker Basket

@Image1 defines the gardener's facial features.

@Image2 defines the gardener's light-green work apron.

@Image3 defines the wicker basket's structure and material.

@Video1 is the original video to extend backward.

Extend @Video1 backward. The first frame of the extended segment must directly follow @Video1's last frame: keep the greenhouse workbench, the gardener's stance and the wicker basket's position continuous.

Then the gardener lifts the wicker basket with both hands and places it on the middle shelf behind them.

Throughout the extension, keep the gardener's face, apron, greenhouse layout and camera direction continuous.

7.3 Forward Extension

First describe what happens before the original clip starts, then write the original clip's first frame as the explicit end state of the extension. Merely writing "then continue the original clip" may cause the model to introduce subsequent characters or effects too early, or to change the image again after reaching the target state.

@Video1 is the original video to extend forward.

Extend @Video1 forward. Before the original video starts,.

The final frame of the extension must naturally meet @Video1's first frame:,, ; keep,, and consistent with @Video1's first frame.

Throughout the extension, keep,, , and continuous.

The same subject remains a single continuous entity throughout — no duplication or splitting. The character's appearance and the object's part count remain stable.

7.4 Forward Extension with Other References

Declare each asset's role one by one, and specify which assets are used in the forward extension vs. only after the original clip starts. This reduces the risk of later characters, props or effects leaking into the pre-segment.

@Image1 defines 's facial features.

@Image2 defines 's clothing.

@Image3 defines 's structure and material.

@Image4 defines two ' grey workwear.

@Image5 defines 's space and lighting.

@Video1 is the original video to extend forward.

Extend @Video1 forward. Before the original video starts, the curator walks to the workbench, picks up the closed wooden display case and opens its lid.

The final frame of the extension must naturally meet @Video1's first frame: the curator stands in the center of the frame, holding the open wooden display case with both hands; the two exhibition assistants stand behind, one on each side. Keep the vertical front medium shot, workbench position, prep-room background and left-side morning light consistent with @Video1's first frame.

must not appear in the forward extension.

Boundary frames should look visually continuous — not pixel-identical. Volume may shift slightly between the extension and the original. Extending video that was itself generated by Seedance 2.5 usually produces smoother visual and audio continuity. During review, check both the boundary transition and the full extended segment.

7.5 Seed Germination Extension (Real 5-Second Case)

Original asset: Seed Germination Video

  • Seedance 2.5 多模态生视频使用指南\图片和附件\0615 发芽.mp4

Original video: seed germination

  • p>
Building on @Video1, extend the clip by 5 seconds. A bee flies in and lands on the flower. The camera pushes in for a macro close-up of the bee's legs and abdomen covered in golden pollen grains. The bee takes off; the camera tracks it to another flower of the same species. In slow motion, pollen falls from the fuzzy body and lands precisely on the stigma — the pollination moment magnified.

Note: choose MOV as the output format for tighter audio/video continuity.

  • Seedance 2.5 多模态生视频使用指南\图片和附件\seedance25-extend-bee-pollination-5s-720p-mov-20260803-baseline-run01_cgt-20260803215229-28wbd.mov

Extension output: bee pollination 5s

  • p>7.6 Seamless Video Transitions
Seamless transitions generate the connective tissue between two video clips. The prompt must first declare which clip is before and which is after, then specify the trigger action, camera motion, how the frame transforms, the arrival state and the sound handoff.

Pre-clip -> Post-clip -> Trigger action -> Camera motion -> Frame morph -> Arrival state -> Sound

Transition

Writing Focus

Dive or bounce-back

State camera direction, speed change and when to enter the next scene

Character spin

State character pose, rotation direction, and how clothing/background transforms continuously

Foreground occlusion

State when the occluder covers the full screen, and the composition that appears after

Object morph

State the shape, material and morphing process of the before/after objects

Push/pull or focus shift

State camera motion, focus subject and the continuous relationship between before/after spaces

@Video1 is the pre-transition clip. Use its.

@Video2 is the post-transition clip. Use its.

Preserve from @Video1 and @Video2.

In @Video1's tail, triggers the transition via.

Camera ; in the frame gradually transforms into the at @Video2's opening.

At the transition's end, naturally arrive at @Video2's opening composition, keeping continuous.

Sound smoothly transitions from to.

Example: Rainy Street -> Exhibition Hall Skylight

@Video1 is the pre-transition clip, featuring a rainy-night street, red umbrella, slow forward push-in and rain ambience.

@Video2 is the post-transition clip, featuring a circular skylight in an exhibition hall, upward tilt and quiet indoor reverb.

Preserve characters, street, hall structure and main action from both originals.

In @Video1's tail, the red umbrella approaches the camera and gradually fills the frame, triggering the transition.

Camera keeps pushing forward; the umbrella's circular edge becomes the hall skylight's metal ring; the red canopy fades into the skylight's white daylight.

At transition's end, naturally arrive at @Video2's opening low-angle composition; camera smoothly switches from forward push to upward tilt.

Rain ambience gradually fades into the hall's footstep echo.

Example: Mahjong Tiles -> Skyscrapers (Real Case)

  • Seedance 2.5 多模态生视频使用指南\图片和附件\1.mp4

Pre-transition: dive view (mahjong)

  • p>
    • Seedance 2.5 多模态生视频使用指南\图片和附件\2.mp4

    Post-transition: dive view (city)

    • p>
    Stitch @Video1 and @Video2 together. @Video1's viewpoint rushes upward to the top and bounces back, diving straight down, then seamlessly transitions into @Video2. During the switch, the mahjong tiles morph into skyscrapers.

    The whole scene transforms accordingly. Do not alter the two source videos themselves.

    • Seedance 2.5 多模态生视频使用指南\图片和附件\积木转场视频.mov

    Seamless transition output: mahjong tiles to skyscrapers

    • p>Chapter 8 — Long-Form Narrative and One-Click Reels8.1 30-Second Long Video: Stages with Explicit End States
    When the story is dense, break it into consecutive stages. Each stage has one main state change, and you must describe the visibly observable state at the end of that stage.

    【Generation Goal】

    Generate a. Core subject is. Main event is.

    【Stage 1】

    Start:.

    Main event:.

    End:.

    【Stage 2】

    Carry-over from previous stage:.

    Main event:.

    End:.

    【Stage 3】

    Main event:.

    End:.

    【Stay Consistent】

    Keep stable.

    Example: Florist Order Wrapping Flow

    【Generation Goal】

    Generate a florist order-wrapping flow demo. The florist and the clerk complete bouquet prep, wrapping and handoff together.

    【Stage 1】

    Start: the florist stands behind the workbench. Loose stems, scissors and wrapping paper sit on the table.

    Main event: the florist arranges stems and trims them.

    End: the bouquet is held in the florist's left hand; the scissors are placed back on the right side of the table.

    【Stage 2】

    Carry-over: both keep the same identity and clothing; the bouquet remains in the florist's hand.

    Main event: the clerk unfolds wrapping paper; the florist places the bouquet in and ties a green ribbon.

    End: the wrapped bouquet lies flat at the table's center, the ribbon knot facing the camera.

    【Stage 3】

    Main event: the clerk picks up the bouquet and places it on the pickup shelf.

    End: the bouquet sits alone at the pickup shelf center; the two stand behind the workbench inspecting the finished piece.

    【Stay Consistent】

    Keep the florist's and clerk's identity, clothing, workbench orientation, scissors position and bouquet ownership stable.

    8.2 Timestamps and Rhythm Control

    For ordinary narrative, prefer "stages". Only use 1-second timestamp expressions for key handoffs, entrances/exits, transitions or explicit beats.

    Format

    Example

    Time range

    `0–3s... 3–7s... 7–12s...`

    Explicit time point

    At 5s, the camera quickly whips left and completes the transition.

    Relative time

    3 seconds after the character presses the button, the room lights gradually dim.

    0–5s: Show an empty wooden display stand; a hand places a white ceramic plate; at end the hand has left and only the white plate remains in the center.

    5–10s: Remove the white plate, place a transparent glass; at end only the transparent glass remains.

    10–15s: Remove the transparent glass, place a green ceramic vase; at end only the green vase remains.

    Time ranges must be continuous and non-overlapping. A time range is a budget for an event, not a precise edit point. Too little content in a range gives the model more room to improvise; too much can cause over-cutting or missed story beats. Do not use timestamps to control frequencies like "shake head three times in one second".

    8.3 Multi-Asset Composition: Tell the Model Which Subject Goes Where

    Seedance 2.5 combines images, videos and audio. With many assets, the prompt's job is not to mention them all in one breath, but to define the mapping between characters, props, scenes, actions and sound. Recommended ordering:

    Per-asset role -> Subject mapping -> Group by type -> Subject definitions -> Per-scene calls

    Step 1: Name each subject individually

    corresponds to @Image1. Adopt only appearance, hair and clothing.

    corresponds to @Image2. Adopt only appearance, hair and clothing.

    corresponds to @Image3. Adopt only structure, material and color.

    references @Image4. Adopt only spatial layout, architecture and light. Do not adopt any character from this image.

    Step 2: Group assets by type

    【Characters】

    The restorer corresponds to @Image1. Adopt only appearance, hair and clothing.

    The recorder corresponds to @Image2. Adopt only appearance, hair and clothing.

    The installer corresponds to @Image3. Adopt only appearance, hair and clothing.

    The docent corresponds to @Image4. Adopt only appearance, hair and clothing.

    The four characters' appearance, clothing, action, blocking and dialogue do not swap between them.

    【Props】

    The sample box corresponds to @Image5 and belongs only to the restorer.

    The recording board corresponds to @Image6 and belongs only to the recorder.

    【Scenes】

    The restoration room references @Image7. Adopt only space, material and light.

    The exhibition hall references @Image8. Adopt only space, material and light.

    【Action and Sound】

    @Video1 supplies the restorer's box-opening action. Do not adopt its characters or scene.

    @Audio1 supplies the docent's voice timbre and assigned dialogue.

    Step 3: Consolidate important cross-scene subjects

    【Subject Definition: Restorer】

    Appearance and clothing: @Image1.

    Fixed prop: the sample box from @Image5.

    Appears in: the restoration room and the exhibition hall.

    Action references: box-opening from @Video1, sample placement from @Video2.

    Excludes: clothing from other characters; never holds the recording board or docent equipment.

    Step 4: Call assets per scene

    Scene One - Restoration Room Inspection

    Uses: restorer, sample box, restoration room and the box-opening action from @Video1.

    Event: the restorer opens the sample box at the workbench and inspects its contents.

    End: the restorer stops inside the workbench; the sample box remains by the restorer's right hand — i.e. on the left side of the frame.

    Scene Two - Exhibition Hall Registration

    Uses: recorder, recording board and exhibition hall.

    Event: the recorder checks the numbers on the recording board beside the display case.

    End: the recorder keeps holding the board with both hands; no other character enters the display case area.

    The goal of multi-asset composition is to let the model pick the correct assets for each scene — not to require every asset to appear simultaneously.

    8.4 One-Click Reel

    One-click reel packages multiple images (or images plus a style reference video) into a coherent video with unified rhythm and packaging. The prompt must describe each asset's role, image order, motion amplitude, edit rhythm, visual packaging and sound. Don't just write "make a video from these assets".

    Asset roles -> Image order -> Motion amplitude -> Edit style -> Visual packaging -> Sound

    【Asset Roles】

    @Image1 is used for.

    @Image2 is used for.

    @Image3 is used for.

    @Video1 is used only for. Do not adopt its characters or scene (optional).

    【Arrangement】

    Images appear in.

    .

    【Frame Dynamics】

    Each image uses.

    Keep stable.

    【Reel Style】

    Adopt.

    【Sound】

    Includes.

    Example: Six-Image Night-Market Travel Reel

    【Asset Roles】

    @Image1 is used for the night-market entrance and opening environment.

    @Image2 is used for the walking along the street.

    @Image3 is used for lantern stalls and craft details.

    @Image4 is used for the three friends dining around a table.

    @Image5 is used for the riverside night view and reflections.

    @Image6 is used for the closing frame of the three friends posing by the bridge.

    @Video1 is used only for the lively edit rhythm, hand-drawn stickers and transitions. Do not adopt its characters or locations.

    【Arrangement】

    The frames appear in the order of @Image1 through @Image6, forming the full arc of "arriving at the night market, strolling, dining, walking, taking a group photo".

    Keep the three friends' appearance and clothing stable; do not swap them.

    【Frame Dynamics】

    Environmental images use slow push-in and slight parallax; character images only add natural blinks, head turns, toasts and clothes swaying in the wind.

    Keep stall structure, table position and bridge railing stable.

    【Reel Style】

    Adopt a brisk travel-vlog rhythm; connect scenes with natural occlusion and harmonious color palettes; hand-drawn stickers appear only at the frame edges.

    【Sound】

    Keep night-market chatter, light cutlery clinks and riverside breeze, paired with brisk but unobtrusive instrumental music.

    8.5 Real Case: Guardian Robot x Rocket Launch (30s Cinematic)

    This example combines multi-grid storyboard, subject references, environment references, dialogue, timestamps and photography terminology.

    image1: nine-cell storyboard reference, supplying the overall shot structure, shot sizes and camera rhythm.

    image2: rocket launch site dusk-grassland real-shot reference, supplying environmental composition, warm gold sunset and cool dusk blue photographic palette.

    image3: Subject 1 (guardian robot) appearance reference.

    image4: Subject 2 (grandmother) appearance reference.

    【Subject Definitions】

    Subject 1 (Guardian Robot): referencing image3, a near-future distressed retro robot with distressed blue-green metal body, mottled rust, dome head, two glowing red round machine eyes, slim antenna and slender jointed limbs. Tall — about twice human height.

    Subject 2 (Grandmother): referencing image4, a petite elderly woman, silver hair in a low bun, deep wrinkles, bright golden floor-length dress with gold-blue embroidered bodice, reluctant expression; reaches only the robot's chest.

    Environment (Dusk Grassland · Launch Site): referencing image2, near-future dusk grassland, sky shifting from warm gold to cool blue, distant launch pad with a white rocket and rising vapor, knee-high grass swaying in the wind, vast and open.

    【Overall Style】

    Live-action photographic color feature film. Photo-real texture throughout, color 35mm film grain, rich cinematic grading, IMAX large-format feel; handheld breathing shake, shallow DOF wide aperture, persistent drifting grass leaves, sparks and ash in foreground, slight Dutch tilt, strong contrast between warm gold sunset, cool dusk blue and explosion warm orange. 16:9 widescreen. Near-future warm-disaster tone: quiet, solemn, protective and reluctant.

    【Strict Exclusions】

    Black-and-white, monochrome, grayscale, desaturated; hand-drawn, sketches, line art, illustration, comic, animation; storyboard / sketch drafts; tilt-shift miniature, doll-like, plastic CGI, greasy overexposed CGI.

    【Shots】(shot structure references image1 nine-cell, 9 shots ≈ 30s)

    Shot 1 (0–3s): Extreme wide, super-low angle flat on the ground looking up, slow handheld dip. Following image2's grassland composition, the dusk grassland stretches vast, foreground knee-high grass blurs and sways with warm gold flare. Two tiny figures sit in the lower third — robot supporting the yellow-dress grandmother, backs to camera, looking toward the distant white rocket venting vapor on the horizon. The massive dusk sky fills most of the frame. SFX: wind, grass rustle, distant engine rumble rising.

    Shot 2 (3–6s): Front medium shot, eye-level with slight handheld breathing. Subject 1 (robot) red eyes dimly glowing, metal hand lightly touching Subject 2 (grandmother)'s arm, reaching only her chest; wind stirs her yellow dress and silver hair; both gaze up at the sky. Shallow DOF, blurred sparks drifting in the foreground. SFX: wind, rising engine rumble. Dialogue (robot, deep warm, English): "I'm right here. I won't let go."

    Shot 3 (6–10s): Face close-up, low-angle slight tilt, handheld breath. Subject 2 (grandmother) wrinkles gilded by sunset, eyes full of reluctance and hope, lips softly murmuring, sad smile forming, realistic skin texture and tear highlights visible, soft flare at edges. Dialogue (grandmother, choked whisper, English): "Fly safe, my child. Come back to me."

    Shot 4 (10–14s): Extreme wide, low angle slow tilt up, handheld. Two tiny figures at the frame's bottom edge looking up, distant horizon white rocket ignites and lifts, flame tears the dusk sky, trailing thick white smoke straight up. Far between characters and rocket; camera follows rocket with unstable upward tilt, foreground grass and flare sweep past, smoke trail motion blur. SFX: rumble peaks then gets pulled back and muffled. Dialogue (grandmother, breathless hope, English): "There he goes... there he goes."

    Shot 5 (14–18s): Extreme wide, camera jolts hard with the shockwave, extremely unstable handheld. Distant high-altitude rocket explodes violently mid-air, body bursting into flying wreckage and thick black smoke — not fireworks, but a tragic, brutal sight; warm orange blast light, blinding foreground flare and scattered sparks, sudden film-grain surge, foreground two figures frozen helplessly looking up. SFX: one muffled explosion then sudden silence. Dialogue (grandmother, gasping, faint, English): "No... no, no—"

    Shot 6 (18–22s): Face extreme close-up, shaking camera slowly stabilizes into uneasy stillness. Subject 2 (grandmother) pupils shrink, an expression of empty disbelief freezes for a beat, then a tear slowly rolls down her wrinkled cheek, lower lip trembles, residual cold-warm light reflected in her tears, shallow DOF, fine film grain. Dialogue (grandmother, hollow disbelief, English): "...he was almost there."

    Shot 7 (22–25s): Close-up to medium shot, handheld trembles with sobs. Subject 2 (grandmother) collapses in full crying, mouth open but silent, trembling hands clutching her chest, tears streaming, shoulders heaving violently with sobs, yellow dress covered in residual ash, foreground flare and ash drifting. Dialogue (grandmother, broken sobbing, English): "Bring him back! Please— bring him back!"

    Shot 8 (25–28s): Super-low angle near vertical tilt-up, camera jolts violently with falling debris. Bottom foreground blurred grass, Subject 1 (robot) bends and wraps its slender arms entirely around Subject 2 (grandmother), shielding her like a protective dome; burning debris falls like fiery rain from the dark sky trailing orange-red tails, robot's back spattered with sparks and scorching flare, debris heavy motion blur, extremely unstable handheld. SFX: whistling debris, metallic thuds. Dialogue (robot, protective resolve, English): "Don't look up. I've got you."

    Shot 9 (28–30s): Wide back shot, super-low angle flat on the ground, shake gradually steadies into a slow pull-back. Foreground blurred grass and warm gold residual flare, Subject 1 (robot) and Subject 2 (grandmother) cling tightly in the dim grassland, she nestles into its slender arms; horizon retains thin smoke, embers fade like fireflies, the last strand of dusk light freezes them into a tender silhouette; camera slowly pulls back and holds. SFX: wind returns, a single light piano note closes the scene. Dialogue (robot, gentle, English): "I'm still here. I'll stay... as long as you need."

    • Seedance 2.5 多模态生视频使用指南\图片和附件\守护机器人_火箭发射_30s.mp4

    30s cinematic - Guardian Robot x Rocket Launch

    • p>Chapter 9 — Advanced Techniques: Camera, Lighting, Expression & Ad Templates9.1 Photography Term Cheat Sheet
    Basic cinematic language and popular camera moves can be written directly into prompts. When a term is uncommon, ambiguous or requires precise visual control, specify who it applies to, how the frame changes and what you want to see.

    Category

    Supported Common Terms

    Shot size

    Extreme wide / wide / medium / close-up / extreme close-up

    Camera motion

    Push / pull / pan / track / dolly / orbit / dive / pull-back / tilt-up / handheld shake

    Camera position & angle

    Low angle / overhead / first-person

    Camera Technique

    Writing Tip

    One-take continuous

    State the subjects, spaces and event order the camera passes through continuously

    Hitchcock zoom

    State the subject's maintained size and how the background is pulled in or pushed away

    Aerial view

    State overhead altitude, motion direction and the environment to show

    FPV

    State first-person flight or traversal path, speed and turn direction

    Bullet time

    State the frozen or slowed action and the camera's orbit direction

    Handheld

    State the follow target and shake degree; avoid only "handheld feel" with no clear camera target

    Speed ramp

    State the action's acceleration, deceleration or rebound point and final rest state

    9.2 Photography Expression Examples

    Aperture, focal length and shutter values can be written into prompts, but the visible result on screen is usually more specific than any single number:

    Example 1. Shallow-DOF portrait: the baker's eyes and face stay sharp; background jars and lights blur into soft round bokeh.

    Example 2. Tracking shot: camera moves horizontally at the same speed as the skateboarder; subject stays sharp, street walls form a right-to-left horizontal motion blur.

    Example 3. Golden hour: warm low-angle sunlight enters from the climber's rear left; mountain ridge casts long shadows.

    Example 4. Natural vignette: the four corners gradually darken; the central pianist's brightness and skin tone remain normal, no black border appears.

    Example 5. Whip-pan transition: at 5s the camera quickly whips left; switch to the next scene when the foreground bookshelf fully occludes the screen, and after the switch the camera keeps moving left at a similar speed.

    9.3 Emotion Direction and Observable Performance

    Just writing emotion words like "tense", "warm", "oppressive" lets the model grasp the overall direction, but it leaves more room for interpretation in performance. To stabilize character acting, describe eyes, brows, mouth corners, breathing, gaze and hand actions — things you can actually see or hear.

    A single emotional turn usually needs only 2–4 clearly observable cues. Only when the emotion has multiple turns do you need to break it into trigger-event stages.

    The overall emotion shifts from to.

    After, the subject first shows.

    Then gradually undergo.

    Finally, the subject expresses through.

    When they hear or see, the subject shows.

    When appears, the subject's expression, gaze or breathing changes.

    After confirming, the subject's attempt to gradually surfaces through.

    Finally, the subject's.

    Example: A young actor's restrained smile after the curtain call

    Applause for the end of the show drifts from backstage. The young actor's fingers clutching the program freeze; her gaze slowly turns toward the curtain; her shoulders stay tense.

    After confirming the bow, she lets out a soft breath, shoulders gradually relax, a restrained smile forms at her mouth corners, her eyes slowly well up — but she never turns to leave.

    9.4 Action and Expression Description Principles

    Two important lessons from S3:

    • Action: prefer broad descriptions (e.g. "did several sets of high-knees and flips", "the two sides engaged in close combat"). Only spell out specific details for the few memorable beats. Do not repeat the same action.
    • Expression: use descriptive sentences; avoid Chinese idioms. For example, instead of "津津有味地吃饭" (eating with great relish), write "face shows a satisfied smile, eating with big bites".
    Chapter 10 — 30-Second Long Narrative and Timeline Extension

    Seedance 2.5 natively outputs 4–30s single clips at a fixed 24fps cinematic frame rate, exporting both MP4 and MOV. Supports 21:9 / 16:9 / 4:3 / 1:1 / 3:4 / 9:16 mainstream aspect ratios.

    10.1 30-Second Living-Room Sofa Commercial (Real Output)

    Use @Video1 as base footage, keep the sofa design, color palette, lighting and camera motion identical.

    Seamlessly extend the clip to a cozy living room scene.

    A woman in off-white knit loungewear sits on the sofa, covered with cream throw blanket while reading.

    She gently sets a coffee mug on the side table. A cat jumps onto the sofa, then a child runs in

    and leans against her as she smiles softly.

    Night falls, floor lamp turns on, the child and cat cuddle quietly together.

    Camera slowly pulls back, leaving clean negative space at the end for brand logo display, total runtime 30 seconds.

    • viddo-seedance-2-5\media\1785562198821-Home_Sofa_Commercial_Extension.mp4

    30s living-room sofa commercial (sofa, characters, cat, child, lamp, logo placeholder)

    • p>10.2 5-Second Timeline Extension (Real Output)
    Extend the @Video1 seed germination footage by 5 seconds.

    A bee flies in and lands on the flower.

    Camera pushes in for macro close-up of golden pollen grains covering its legs and abdomen.

    The bee takes flight, camera tracks it to another flower of the same species.

    Slow-motion shot captures pollen falling from fuzzy body and landing precisely on flower stigma to visualize pollination.

    • viddo-seedance-2-5\media\1785562527157-High_Fidelity_Timeline_Extension_1.mp4

    5s timeline extension 1 - bee pollination

    • p>
      • viddo-seedance-2-5\media\1785562531248-High_Fidelity_Timeline_Extension_2.mp4

      5s timeline extension 2 - slow-motion pollen drop

      • p>Chapter 11 — 50 Multi-Modal Assets and @ Reference Syntax
      Viddo.ai's integrated Seedance 2.5 accepts up to 50 mixed reference assets per generation:

      Asset

      Limit

      Size / Duration

      Reference images

      0–30 files

      Each ≤ 30 MB

      Reference videos

      0–10 files

      Each MP4/MOV ≤ 200 MB, 2–30s

      Reference audio

      0–10 files

      Total combined runtime with video ≤ 30s

      After upload, the platform auto-assigns unique IDs (e.g. `@Image1`, `@Video1`, `@Audio1`). Standard practice: follow each ID with a one-line usage description to lock the visual binding.

      Standard correct syntax example

      immersive concept short film, use @Image1 seaside villa as scene reference to lock building architecture;

      slow handheld push-in camera movement, lighting shifts with natural morning daylight.

      11.1 Multi-Asset Combo: House Assembly Animation

      Use @Image1, @Image2, @Image3, @Image4 as structural reference assets,

      generate full assembly animation of the house from base framework to finished build,

      consistent proportions and textures across all frames.

      • viddo-seedance-2-5\media\1785563856645-Exclusive_Reference_Syntax.mp4

      @Image1-4 structural reference - house assembly animation

      • p>11.2 Exclusive: Audio-Only Reference Video Generation
      Seedance 2.5 is the industry's first model with native audio-only-driven visual generation. Upload background music or ambient sound as `@Audio` references and the model auto-matches shot rhythm, edit points and color grading to the audio beats. Perfect for music videos, atmospheric shorts and soundtrack-aligned commercials.

      • viddo-seedance-2-5\media\1785563210229-Audio_Only_Reference_Video_Generation.mp4

      @Audio-driven visuals - automatic beat matching

      • p>Chapter 12 — Precise Local Video Editing (Non-Destructive Refinement)
      Seedance 2.5 preserves the original camera framing, motion and overall rhythm, only modifying the specified visual regions. Four core workflows: partial repaint, subject replacement, global lighting rework, add/remove visual elements.

      12.1 One-Click Visual Style Swap (Locked Camera, Variable Style)

      Keep all camera timing and framing from @Video1 unchanged,

      sequentially switch visual styles:

      American gritty photorealism, Chinese ink wash animation, black-and-white vintage comic, Japanese anime.

      All character proportions and body forms remain identical throughout style shifts.

      • viddo-seedance-2-5\media\1785563664051-One_Click_Visual_Style_Swap_1.mp4

      One-click style swap 1 - American / ink wash / B&W / anime

      • p>
        • viddo-seedance-2-5\media\1785563893641-One_Click_Visual_Style_Swap_2.mp4

        One-click style swap 2 - consistent character proportions

        • p>12.2 Global Scene Lighting Rework
        Rewrite harsh midday overhead lighting from @Image1 reference footage

        into soft sunset side backlight,

        add natural ambient diffuse reflection, realistic layered skin shading for human subjects.

        • viddo-seedance-2-5\media\1785564063569-Global_Scene_Lighting_Rework.mp4

        Global lighting rework - harsh overhead to sunset side backlight

        • p>12.3 Object Removal and Seamless Background Fill
        Remove the drone and edge tracking vehicle in the bottom-left frame,

        fill empty space naturally with matching savanna, tree branches and golden sunset backlight.

        Keep giraffes, distant landscape, sunset rim light and original camera composition fully intact.

        No distortion or unnatural visual artifacts after filling.

        • viddo-seedance-2-5\media\1785564188109-Object_Removal_1.mp4

        Object removal 1 - drone and chase car into natural grassland

        • p>
          • viddo-seedance-2-5\media\1785564194588-Object_Removal_2.mp4

          Object removal 2 - detail recovery and camera stability

          • p>12.4 Character Age Progression and Micro-Expression Editing
          Preserve all camera framing, lighting and performance timing from @Video1.

          Gradually age the female lead from her 20s to 60s;

          shift her reserved sad emotion into gentle warmth,

          tears roll down eye corners,

          she slowly lifts the corners of her mouth and reveals a relieved smile through soft tears.

          • viddo-seedance-2-5\media\1785564316549-Fine_Grained_Character_Age.mp4

          Age / micro-expression edit - 20s to 60s, tears to relieved smile

          • p>Chapter 13 — Visual Quality Upgrade: Lighting, Camera & Facial Expression13.1 Physically Based Realistic Lighting (vs 2.0)
          Seedance 2.5 no longer relies on simple filter-based lighting — it fully simulates real-world light physics: directional falloff, surface bounce, contact shadows and environmental reflection layers.

          • Transparent, natural eye highlights without artificial plastic feel
          • Layered contact shadows, environmental bounce light and reflective surface rendering
          • Consistent light logic for outdoor sunrise/sunset, indoor floor lamps and studio softboxes
          • viddo-seedance-2-5\media\1785564487168-Visual_Quality_Upgrade_1.mp4

          Quality upgrade 1 - physical lighting and skin texture

          • p>
            • viddo-seedance-2-5\media\1785564492270-Visual_Quality_Upgrade_2.mp4

            Quality upgrade 2 - eye highlights and environmental bounce

            • p>13.2 High-Precision Controllable Camera Motion
            Full cinematic camera execution with zero jitter, frame loss or shot drift:

            • Basic moves: push / pull / horizontal pan / vertical tilt / object tracking
            • Compound moves: crane lift / orbital rotation / multi-path follow

            Two young children walk side-by-side across wild highland terrain.

            Distant strange sound echoes, they stop and slowly look upward to the sky with tense, uneasy expressions.

            A shooting star streaks across the sky, camera pans smoothly following its trajectory,

            then pulls back wide to watch the star vanish over the horizon.

            • viddo-seedance-2-5\media\1785564606356-High_Precision_Controllable_Camera_Motion.mp4

            High-precision camera - shooting-star tracking + pull-back

            • p>13.3 Character Performance and Micro-Expression Recognition
            Young East Asian woman sits by study window at dusk,

            slow push-in close-up shot capturing the full emotional shift from teary eyes to soft smile

            as she reads a letter.

            Letter focus gradually blurs in sync with emotional beats.

            • viddo-seedance-2-5\media\1785564693238-Advanced_Character.mp4

            Character micro-expression - tears to soft smile

            • p>Chapter 14 — Commercial Workflows and Prompt Templates14.1 E-commerce and Advertising
            30s 16:9 Furniture Ad Template

            30-second 16:9 sofa advertisement, first 15 seconds feature product close-ups in white studio lighting,

            second half shifts to warm cozy living room environment.

            Cast includes one woman, one child and a cat, soft natural photorealistic lighting.

            Lock sofa product design via @Image1 reference, consistent material and color across all frames,

            clean logo placeholder at the end frame.

            • viddo-seedance-2-5\media\1785564806297-30s_Home_Furniture_Commercial_Template.mp4

            30s furniture ad template - white studio to cozy living room

            • p>
            Reusable SKU Product Reel Template

            Use @Image1 soda beverage as core product reference,

            @Image2 indoor lifestyle scene as environment reference.

            Generate sequential shot list: product macro close-up -> medium handheld shot -> wide lifestyle layout shot.

            Maintain 100% consistent soda packaging across all frames with smooth camera transitions

            for bulk e-commerce listing videos.

            • viddo-seedance-2-5\media\1785564886866-Reusable_SKU_Product_Reel_Template.mp4

            Reusable SKU reel template - drink close-up to lifestyle

            • p>
            3D Animated Snack Commercial Template

            Bright, transparent 3D cartoon visual style,

            starring an expressive desert horned lizard with detailed fur texture and dreamy shallow depth of field.

            Opening scene on hot desert sand, lizard flicks tongue for playful interaction,

            soft natural diffused lighting,

            blending photorealism and childlike whimsy for snack promotional reels.

            • viddo-seedance-2-5\media\1785565127584-3D_Animated_Snack_Commercial_Template.mp4

            3D snack commercial - desert horned lizard

            • p>14.2 Film and Entertainment
              • viddo-seedance-2-5\media\1785566409675-6_2_1.mp4

              30s one-take - busy restaurant kitchen

              • p>
                • viddo-seedance-2-5\media\1785566419062-6_2_2.mp4

                Music video / film pre-vis / age-progression character

                • p>
                  • viddo-seedance-2-5\media\1785566429689-6_2_3.mp4

                  3D gray model to photoreal render - architecture / product

                  • p>14.3 Education Content
                    • viddo-seedance-2-5\media\1785565737560-6_3_1.mp4

                    30s multilingual cultural-heritage museum documentary

                    • p>
                      • viddo-seedance-2-5\media\1785565746065-6_3_2.mp4

                      Stop-motion - currency evolution / biological growth

                      • p>
                        • viddo-seedance-2-5\media\1785566579917-6_3_3.mp4

                        Industrial mechanical demo - swap robotic arms, keep camera motion

                        • p>Chapter 15 — Parameter Cheat Sheet and FAQ15.1 Parameter Cheat Sheet
                        Parameter

                        Seedance 2.5 Specification

                        Aspect ratio

                        21:9 / 16:9 / 4:3 / 1:1 / 3:4 / 9:16; editing/extension tasks inherit original asset ratio automatically

                        Resolution

                        480p, 720p; higher resolution coming in future updates

                        Frame rate

                        Fixed 24fps cinematic frame rate

                        Video duration

                        Integer 4–30 seconds; edited clips inherit source runtime

                        Output format

                        MP4 and MOV dual output

                        Concurrent render limit

                        Personal accounts: 3 concurrent; enterprise commercial accounts: 10 concurrent

                        Render speed

                        Personal: 180 frames per minute; enterprise: 600 frames per minute

                        Max reference assets

                        50 mixed files: 30 images + 10 videos + 10 audio clips

                        15.2 FAQ

                        Q1: How many reference images can I upload to Seedance 2.5?

                        A: Up to 30 reference images per generation, paired with up to 10 reference videos and 10 audio clips — 50 multi-modal assets in total. Audio-only reference mode is supported for sound-driven visual creation.

                        Q2: What is the maximum video length Seedance 2.5 can generate?

                        A: Native single-shot generation supports up to 30 continuous seconds. Short clips can be extended via timeline extension tools for complete TVC and short-drama content.

                        Q3: How do I use the @ reference syntax inside Viddo.ai?

                        A: After uploading media to your asset library, the platform auto-assigns unique IDs like @Image1 or @Video1. Add a sentence explaining each asset's purpose right after the ID to lock consistent visual identity.

                        Q4: Does Seedance 2.5 support non-English prompts?

                        A: The model natively supports 10+ global languages including Chinese, English, Japanese, Korean and Southeast Asian languages. No translation middleware required, with vastly improved complex-prompt comprehension accuracy.

                        Q5: Can I partially edit finished AI-generated videos?

                        A: Yes. Local editing tools support partial repainting, subject replacement, lighting rework and unwanted object removal. Original camera framing, shot timing and scene composition stay fully locked without full re-rendering.

                        Chapter 16 — Pre-Submission Checklist and Usage Limits16.1 Pre-Submission Self-Check
                        • Have you clearly stated the subject and main action or event?
                        • Does each reference asset specify what to adopt and what to exclude?
                        • Are different characters, products and props named individually and bound to assets?
                        • Are multi-assets selected per scene rather than forced to all appear at once?
                        • Does each long-video stage have only one main change with an explicit end state?
                        • Are character count, clothing, prop ownership and spatial relationships stable?
                        • Does video editing specify a single master, edit range, target count and preserved content?
                        • Do abstract emotions and photography terms map to actually observable cues?
                        • Are first/last frames and multi-keyframes documented one by one, with first and last frames using the same aspect ratio?
                        • Does the storyboard clarify inherited structure; does the gray model declare coarse or fine-grained and the timing/structure/material/style to inherit?
                        • Do video editing, first/last-frame generation and video extension follow the auto-locked ratio and duration rules?
                        • Does video extension verify boundary frames, motion trend and sound continuity together?
                        • Does one-click reel specify asset roles, image order, motion amplitude, edit style and sound?
                        • Does a seamless transition declare both clips' roles, trigger action, transition process and arrival state?
                        16.2 Usage Limits (Mandatory Reading)
                        • Timestamps allocate event rhythm only, not frame-accurate edit points.
                        • Video editing prompts can improve alignment probability with the source video but cannot guarantee frame-perfect replication.
                        • Multi-asset composition is about choosing and combining assets correctly, not making every asset appear simultaneously.
                        • Hard-accuracy subtitles, formulas, signage, product parameters and frame-accurate time points should be done through pre-composited assets + video generation + post-production.
                        • Video editing locks the input video's ratio and base duration; both are non-configurable. Output duration may differ by up to ~0.3s from the input.
                        • First-frame or first/last-frame generation locks the ratio to the first frame; duration can be set. Different ratios between first and last frames may stretch the last frame.
                        • Video extension locks the input video's ratio; extension duration can be set. Volume in the extended segment may shift slightly relative to the original.
                        • In one-click reel, if image order or character-to-character mapping matters, you must specify it explicitly in the prompt.
                        • Seamless transitions aim for visual and sound continuity, not pixel-identical preservation of the two source clips.

                        About the Author

                        Share practical tips on AI tools and provide in-depth analysis of their features and use cases to boost productivity in both work and daily life.

                        Rate this Article
Leave a Comment
Author Thumbnail
I Agree:
Comment 
Pictures
Author: Aivon Ai

Aivon Ai

Member since: Sep 03, 2026
Published articles: 3

Related Articles