Multimodal input
Combine text with images, video clips, and audio references to communicate the full creative direction.
使用 Seedance 2.0 創作電影感故事
將文字和多模態參考轉化為連貫的多鏡頭影片，用自然語言控制角色、鏡頭、節奏與聲音。
生成記錄
登入後查看生成記錄。
SEEDANCE 2.0 範例
從創意簡報到完整動態場景
探索 AI 影片工作流能夠呈現的鏡頭、風格、動作和視覺敘事。
ONE MODEL, MORE CONTROL
Build video with references the model can understand
Seedance 2.0 is designed around multimodal direction, letting creators combine a prompt with visual, motion, and audio references.
Consistent characters
Keep faces, clothing, products, scenes, and visual style coherent across connected shots.
Camera and motion control
Reference choreography, camera paths, timing, transitions, and effects instead of describing every detail.
Multi-shot narratives
Plan connected scenes with natural cuts, evolving action, and continuity across a short sequence.
Audio-aware creation
Guide dialogue, ambience, sound effects, and musical rhythm alongside the visuals.
Video extension and refinement
Continue a clip or iterate on selected creative decisions while preserving its direction.
AT A GLANCE
Seedance 2.0 input and output capabilities
Input typesText, images, video, and audio references
Reference imagesUp to 9 images
Video referencesUp to 3 clips, subject to platform limits
Audio referencesUp to 3 clips, subject to platform limits
Clip durationUp to 15 seconds per generation
Creative controlCharacters, motion, camera, transitions, style, and sound
BUILT FOR CREATORS
One AI video workflow for every stage of production
Short films
Develop coherent scenes with recurring characters, deliberate camera work, and connected action.
Video advertising
Turn product imagery and campaign direction into multiple polished ad concepts.
Social video
Rework creative ideas into fast-moving vertical clips for Reels, TikTok, and Shorts.
Music and dance
Use movement or audio references to guide choreography, tempo, and visual rhythm.
Pre-visualization
Test shot composition, camera movement, transitions, and VFX ideas before production.
Education and demos
Create visual explanations, product walkthroughs, and story-led learning content.
HOW TO USE SEEDANCE 2.0
Create an AI video in three steps
Start simple, give every reference a clear role, then refine only the part that needs more control.
Add your references
Upload the images, video, or audio that define the character, first frame, motion, camera, or soundtrack.
Describe the sequence
Write the subject, action, shot order, camera language, lighting, sound, and duration in natural language.
Generate and iterate
Review continuity and timing, then simplify the prompt or strengthen a reference to improve the next result.
PROMPT GUIDE
Write prompts like a compact shot list
A strong Seedance 2.0 prompt explains what happens over time. Name the subject, action, camera, environment, visual style, sound, and transitions.
Subject + timed action + camera + environment + style + sound
A ceramic robot walks through a rain-soaked night market. Start with a wide establishing shot, track beside the robot, then cut to a close-up as neon reflections move across its face. Slow cinematic pacing, soft rain ambience, distant crowd chatter.