
“Before MACE-Dance, I needed dancers, rehearsal time and manual editing for every concept. Now I test music-driven choreography ideas from a prompt and reference image before booking a studio.”
把音乐节奏、动作意图和视觉描述转成舞蹈视频。先在首页设置模型、比例、时长和动作风格,再进入完整视频生成器。
MACE-Dance 创作流程已就绪。

Use text, images, videos, audio, camera motion, aspect ratio and duration to generate a MACE-Dance-style AI video.
音乐驱动 • 文生视频 • 运镜编舞
MACE-Dance 通过 Motion-Appearance Cascaded Experts,将音乐驱动的舞蹈视频生成拆成动作、外观、评估等关键能力。

MACE-Dance 将节奏、拍点、速度变化和音乐情绪映射为与音乐对齐的 3D 舞蹈动作,让舞者跟随歌曲运动,而不是生成普通的视频主体动作。



围绕 AI 舞蹈视频生成器、音乐生成舞蹈视频、文生舞蹈视频和参考图生成舞蹈视频建立清晰入口。

从音乐、文字提示和参考视觉创建可用于编舞预览的舞蹈视频,适合全身动作、稳定身份和音乐对齐表演。
立即试用 →
把歌曲、节拍或音乐情绪转成同步舞蹈动作和表演。Motion Expert 会读取音乐结构,让步伐、停顿和能量变化匹配音频。
立即试用 →
描述舞者、舞台灯光、镜头运动和视觉风格,生成 K-pop、Hip-hop、偶像舞台、电影运镜或竖屏舞蹈视频概念。
立即试用 →
用角色或舞者参考图约束外观,同时让生成动作跟随音乐,适合虚拟偶像、头像动画、品牌角色和身份保持舞蹈视频。
立即试用 →
生成用于全身人物舞蹈动画和动作评估的 SMPL 风格动作线索,将 3D dance motion、运动学合理性和 MA-Data 转为实际创作能力。
立即试用 →
用音乐驱动编舞让虚拟偶像、风格化角色和表演者参考图动起来,并在社交媒体、音乐视频和舞台预演中保持角色可识别。
立即试用 →覆盖 K-pop 舞蹈、虚拟偶像、社交媒体、音乐视频和动画预演等长尾搜索场景。

生成高能 K-pop 舞蹈片段,用于短视频、偶像概念和舞台预演,提前测试编舞、节奏重音、镜头运动和视觉风格。

在正式制作前快速预演音乐视频编舞、镜头运动和视觉方向,减少每个创意都需要舞者或动捕排期的成本。

用音乐对齐的全身舞蹈动作驱动虚拟偶像或角色参考图,保持身份、服装和风格稳定。

为 TikTok、Instagram Reels、YouTube Shorts 和趋势测试创建可分享舞蹈视频,支持竖屏舞蹈和快速内容迭代。

在游戏、动画和虚拟演唱会中预览舞蹈动作,让团队先评估全身运动、姿态可读性、节奏和角色表现。

用动作-外观评估关键词衡量输出质量,将 MA-Data、VBench 和 ViTPose 关键点与真实视频审查连接起来。
创作者用 MACE-Dance 替代缓慢的编舞预览、昂贵的动捕规划和手工动画草稿,快速生成 AI 舞蹈视频。

“Before MACE-Dance, I needed dancers, rehearsal time and manual editing for every concept. Now I test music-driven choreography ideas from a prompt and reference image before booking a studio.”

“The Motion Expert and Appearance Expert workflow fits virtual idol production because the dance follows the beat while the character identity stays consistent across the generated video.”

“We use MACE-Dance to draft K-pop dance videos, TikTok dance clips and short music video concepts. It is faster than waiting for animation previews or expensive mocap sessions.”

“For previs, the important part is whether the motion reads clearly before the team spends time on final animation. MACE-Dance gives us beat-aligned dance motion and a coherent visual target in minutes.”

“The separation between music-to-3D dance motion and appearance-conditioned video synthesis makes the product story much clearer than a generic text to video AI workflow.”

“Our team tests vertical dance ideas, branded character moves and virtual performer clips with MACE-Dance before deciding which concepts deserve production budget.”
Choose the plan that works best for you. All plans include access to our core features.
适合个人用户和轻度使用
适合专业创作者和团队
面向大型企业和专业工作室
One-time purchase • Credits never expire • For individuals and teams
适合体验服务和小型项目
适合常规用户,超值之选
适合专业人士和企业用户
适合重度用户和团队使用
Search-focused answers for music-driven dance video generation.
MACE-Dance is a Motion-Appearance Cascaded Experts framework for music-driven dance video generation. It produces realistic full-body dance motion first, then synthesizes a visually coherent dance video with stable appearance, character identity and camera-ready output.
Describe the dancer, music mood, choreography style, camera direction, lighting and output format. MACE-Dance creates beat-aligned 3D dance movement, then applies motion-conditioned video synthesis to render the final AI dance video.
The Motion Expert focuses on music-aligned 3D dance motion, kinematic plausibility and artistic expressiveness. The Appearance Expert converts that motion into coherent video with stable identity, reference-conditioned appearance and visual style.
Yes. MACE-Dance supports K-pop dance generation, hip-hop choreography, virtual idol dance, social media dance clips and cinematic music videos. Users can specify tempo, stage mood, camera movement, dancer style and platform format.
Dance video generation needs stable identity, consistent body appearance and smooth motion over time. Spatiotemporal coherence keeps the dancer believable frame by frame, reducing flicker, body distortion, outfit drift and unstable backgrounds.
Generic AI video generators often treat dance as one visual prompt. MACE-Dance separates motion planning and appearance synthesis, so the system can focus on music alignment, full-body choreography, identity preservation and high-fidelity human animation.
Yes. A reference image can guide dancer or character appearance while the generated motion follows the music. This supports virtual idols, avatars, branded characters, stylized performers and identity-preserving dance videos.
The free plan is designed for testing AI dance video workflows, prompt structure, music-driven choreography previews and reference image to dance video experiments before upgrading to higher quality or commercial usage.
Commercial usage depends on the selected plan. Paid creator, pro and studio plans are intended for music videos, social media campaigns, virtual idol production and branded dance content.
Generated outputs are intended for the creator account that produces them, subject to plan terms, uploaded source rights and commercial license conditions for music, character references and final dance video usage.
从歌曲、舞者参考图或编舞提示开始,在一个工作流中创建 AI 舞蹈视频、K-pop 舞蹈、虚拟偶像动作和电影感音乐视频概念。