AI Video Generation
Transform your ideas into stunning videos with AI. Choose from multiple high-quality models optimized for different use cases.
Example Prompts:
- • A cute cat playing with butterflies in a sunny garden
- • City skyline at sunset with golden light reflecting off skyscrapers
- • Ocean waves crashing on a beach during golden hour
- • A person walking through a cherry blossom forest in spring
🎬
视频生成预览
填写表单后点击「生成视频」开始
内容政策提示
不符合政策的内容生成将失败——这不是诈骗或网站错误。Kling 4 模型最近收紧了内容限制，所有真人面部输入（自拍、肖像、名人）现在在模型层面被阻止。暴力和 NSFW 内容也会被拒绝。您仍可以使用插画、动漫角色、AI 生成的面部或风格化艺术作品创作精彩的视频。
Image-to-Video Modes
In addition to text-to-video, we also support various image-to-video modes for more creative freedom.
🎬
First Frame Generation
Upload 1 image as the first frame of the video
• Supported: All models
• Image count: 1
🎨
Reference Image GenerationDoubao
Upload 1-3 images as style references
• Supported: Doubao Seedance
• Image count: 1-3
🔄
Start-End Frame Generation
Upload 2 images to generate the transition in between
• Supported: Doubao, MiniMax
• Image count: 2
Model Comparison
Veo 3.1 (Google)
Fixed 8-second duration, high quality, fast generation
- • 16:9 or 9:16 aspect ratio
- • ~2 minutes processing
- • 260-520 credits
Doubao Seedance (ByteDance)
Flexible 2-12 second duration, most affordable
- • 6 aspect ratios available
- • Reference image support
- • 200-520 credits
MiniMax Hailuo (MiniMax)
5 or 10 second duration with prompt optimization
- • Up to 1080p resolution
- • Auto prompt optimization
- • 245-370 credits