動画を作成: Seedance 2.0
Seedance 2.0 is a video model for cinematic motion, multi-shot consistency, image-guided video, and reference-driven creation. Use it for ad concepts, character action, scene transitions, and social clips.
How this model fits creative work
Seedance 2.0 is a video model for cinematic motion, multi-shot consistency, image-guided video, and reference-driven creation. Use it for ad concepts, character action, scene transitions, and social clips.
テキストから動画
Seedance 2.0 is a video model for cinematic motion, multi-shot consistency, image-guided video, and reference-driven creation. Use it for ad concepts, character action, scene transitions, and social clips.
画像から動画
Generate short variations before production to compare camera motion, action, pacing, and subject behavior.
参照から動画
需要用多张图片约束主体、风格、道具或构图时，使用参考生视频。参考图能强化视觉方向，但仍需要明确文字说明。
このモデルが向いている用途
Seedance 2.0 is a video model for cinematic motion, multi-shot consistency, image-guided video, and reference-driven creation. Use it for ad concepts, character action, scene transitions, and social clips.
Model controls and capabilities
Seedance 2.0 shows only the controls currently supported in the generator, so creators can focus on prompts, references, and output settings.
Fast mode: 480p and 720p
Choose the output level that fits draft review or more detailed visual checks.
Quality mode: 480p, 720p, and 1080p
Choose a duration that fits the motion idea, then refine after checking action and camera behavior.
4 to 15 second duration
Choose landscape, portrait, or square framing for ads, landing pages, and social platforms.
16:9, 4:3, 1:1, 3:4, 9:16, and 21:9 frames
Use a seed to create similar variations while comparing prompt and reference changes.
Text, first/last frame, and reference image workflows
Use text, first/last frames, or reference images depending on the creative goal.
より良い結果を得る方法
把 Seedance 2.0 任务当成短视频制作 brief，而不是一个短关键词。动作和视觉约束越明确，结果越容易判断。
选择正确模式
新想法用文生视频，首尾帧控制用图生视频，多参考约束用参考生视频。
把动作写进提示词
写清什么在动、镜头如何运动、画面如何变化，以及最后一帧应是什么状态。
Test motion with short clips
Use early generations to check action, camera logic, subject consistency, and pacing before refining the output.
一次只改一个变量
先检查主体一致性、物理运动、镜头逻辑和节奏，再逐步调整。
Frequently asked questions
Guidance for Seedance 2.0 inputs, use cases, and output controls.
Seedance 2.0 is a video model for cinematic motion, multi-shot consistency, image-guided video, and reference-driven creation. Use it for ad concepts, character action, scene transitions, and social clips.
動画生成を始める with Seedance 2.0
Open the generator, choose this model, and submit a clear prompt or reference image for a short video task.