为什么选择 mawu.ai
图生视频 – 前所未有地简单
一条出色短视频所需的一切,尽在一个工作室。
- 以你的图片为起点照片就是片段的第一帧,人脸、产品和色彩都与原图保持一致。
- 首帧和尾帧再指定最后一帧,AI会自动补全中间的动态。
- 有声Seedance 2.5会同步生成与画面相配的声响和氛围。
- 适配各种格式9:16适合TikTok和Reels,16:9适合YouTube,1:1适合信息流——支持480p、720p或1080p,使用Gemini Omni 1.2 Flash甚至可达4K。
使用方法
三步完成视频
- 1
上传图片
一张照片、一张产品图,或媒体库中的任意图片。
- 2
描述动态
例如:镜头缓缓推近,人物微笑着转过身来。
- 3
创建片段
下载成片,或在编辑器中继续剪辑。
模板
从现成模板开始
全部出自mawu的真实成果,每一件都附上当时所用的提示词——一键即可开始创作。
金色之夜
片段“Cinematic widescreen night shot, photoreal, 1980s small town summer, 35 mm anamorphic film look with soft grain. Six teenagers in retro 80s clothes, striped t-shirts, denim and sneakers, seen from behind and in side silhouette, stand on a dark grassy hill; two of them hold hands, one points up at the sky. Huge golden-yellow fireworks burst across the black sky and slowly rain down glittering gold sparks. Far below, a small town glows in warm yellow light under low golden smoke. Yellow and gold are the only accent colours, deep blue-black shadows, no red anywhere. Slow steady camera push-in towards the group, gentle wind in the grass, fireworks flashing light over their backs. Soft distant firework booms and crickets. No text, no logos, no titles.”
Car Edit
素材库武士大战巨龙
片段“Cinematic 35 mm film shot, vertical. A lone samurai in rain-soaked black armor stands on a storm-lashed cliff at night while a colossal ancient dragon rises out of the black ocean behind him, lightning tearing the sky apart and huge waves exploding against the rocks. The camera rushes forward and whips around him as he slowly draws a glowing katana, looks straight into the lens and shouts: "Not today." Deafening thunder, roaring waves, the dragon's low growl, rising orchestral percussion. Hyper-detailed, dramatic rim light. No text, no subtitles, no logos.”
30秒看日本
素材库武士预告片
片段“Cinematic 35 mm film shot, vertical. A lone samurai in rain-soaked black armor stands on a storm-lashed cliff at night while a colossal ancient dragon rises out of the black ocean behind him, lightning tearing the sky apart and huge waves exploding against the rocks. The camera rushes forward and whips around him as he slowly draws a glowing katana, looks straight into the lens and shouts: "Not today." Deafening thunder, roaring waves, the dragon's low growl, rising orchestral percussion. Hyper-detailed, dramatic rim light. No text, no subtitles, no logos.”
Gym Edit
素材库肌肉网红热舞
片段“Funny viral trend clip: an over-the-top fitness influencer with absurdly huge cartoonish muscles, a very long neck and a long braided beard dances with full confidence in a sunny gym, doing a goofy hip-hop dance with chest bounces and a spin, perfectly on the beat, gym members in the background laugh and film him on their phones. Photorealistic but exaggerated proportions, bright light, handheld phone camera that bounces along. Sound: a catchy upbeat hip-hop beat with a heavy bass drop, sneakers squeaking, people laughing. No text, no logos, unbranded clothes.”
自律胜过激励
素材库
常见问题
常见问题
哪个AI模型来为我的图片添加动画?
ByteDance的Seedance 2.5和2.0、Kuaishou的Kling 4.0、Google的Gemini Omni 1.2 Flash以及MiniMax H3。Seedance 2.5和Kling 4.0可生成最长30秒的有声视频,Gemini Omni 1.2 Flash最高可达4K,Seedance的Fast和Mini版本价格更低。
我的图片还能保持原样吗?
可以。你的图片就是片段的第一帧,AI从这里开始让它动起来。加上结束帧,还能决定动态的落点。
片段有多长?
视模型而定,最长12、15或30秒。更长的视频可在编辑器中用多个片段拼接而成。
一个片段要多少钱?
片段消耗的积分取决于模型、时长和分辨率。创建前即可看到确切价格,只有成功生成的片段才会扣费。