Kling 3.0 Pro 画像から動画生成可将单张入力图像制作成高质量、逼真的视频,并加入流畅的镜头运动、自然物理效果和出色的时间一致性。它擅长处理真实场景、人物运动、环境细节和シネマティック运动,同时保留原始图像的构图与光照。
このモデルについて
Kling 3.0 Pro 画像から動画生成是一项前沿解决方案,可将单张静态图像转化为无缝、高质量视频。借助先进的 AI 技术和自然语言处理能力,它能够生成逼真的镜头运动、流畅的转场和动态环境细节。该技术擅长保持原始图像的结构与光照,确保生成的每个片段都兼具视觉吸引力并忠于源素材。
凭借出色的时间一致性和自然物理效果,该模型能够呈现捕捉真实场景与人物运动本质的シネマティック运动。无论是普通照片还是细致的环境场景,Kling 3.0 Pro 都能为希望通过动态视频增强数字叙事的内容创作者提供可靠、高效且直观的工作流。
コスト分析
| プロバイダー | 費用 | 備考 |
|---|---|---|
| muapiapp | $0.72 per generation | muapiapp is 20-50% more affordable than its competitors, delivering comparable or superior quality at a lower cost. |
| Fal.ai | $0.90 per generation | Compared to Fal.ai, muapiapp offers a 20-50% cost saving while maintaining high-quality output and performance. |
| Replicate | $0.90 per generation | muapiapp is 20-50% cheaper than Replicate, providing an economical solution without compromising on professional-grade results. |
muapiapp is 20-50% more affordable than its competitors, delivering comparable or superior quality at a lower cost.
Compared to Fal.ai, muapiapp offers a 20-50% cost saving while maintaining high-quality output and performance.
muapiapp is 20-50% cheaper than Replicate, providing an economical solution without compromising on professional-grade results.
** 競合サービスの料金は類似のモデル構成および利用ティアに基づいて算出された推定値です。
設定スキーマ
| パラメータ | 型 | 説明 | デフォルト |
|---|---|---|---|
| プロンプト | string | 描述動画内容的文本プロンプト。 | The camera begins on the railway station platform beside a stationary train as morning sunlight filters through the roof. Passengers make small natural movements while the train doors are open. The camera moves forward and enters the train, transitioning smoothly into a window-seat point of view. As the doors close, the train starts moving. The view shifts fully to the window, showing the city passing by outside with gentle motion blur, buildings and trees sliding past. Sunlight reflects on the glass, faint interior reflections appear, and the ride feels calm and realistic with smooth, cinematic motion. |
| 画像 URL | string | 用于生成動画的入力画像 URL。 | https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/kling-v3.0-pro-image-to-video1.jpg |
| 最后一帧画像 | string | 入力末帧画像的 URL。 | https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/kling-v3.0-pro-image-to-video2.jpg |
| 長さ(秒) | int | 生成動画的生成時間(秒)。 | 5 |
| 生成音声 | boolean | 是否为動画生成音声 | true |
描述動画内容的文本プロンプト。
The camera begins on the railway station platform beside a stationary train as morning sunlight filters through the roof. Passengers make small natural movements while the train doors are open. The camera moves forward and enters the train, transitioning smoothly into a window-seat point of view. As the doors close, the train starts moving. The view shifts fully to the window, showing the city passing by outside with gentle motion blur, buildings and trees sliding past. Sunlight reflects on the glass, faint interior reflections appear, and the ride feels calm and realistic with smooth, cinematic motion.用于生成動画的入力画像 URL。
https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/kling-v3.0-pro-image-to-video1.jpg入力末帧画像的 URL。
https://d3adwkbyhxyrtq.cloudfront.net/webassets/videomodels/kling-v3.0-pro-image-to-video2.jpg生成動画的生成時間(秒)。
5是否为動画生成音声
true開発者ドキュメント
準備入力
リクエスト送信
prompt 和 image_url。查看和結果の確認
FAQ
高分辨率、主体清晰且光照良好的图像效果最佳。具有明确元素和清晰结构细节的图像,有助于 AI 创建更准确、更具吸引力的动画。
你可以使用 ‘duration’ 参数将视频时长设置在 3 到 15 秒之间。此外,‘generate_audio’ 布尔参数允许你选择生成的视频是否包含音轨。
会。该模型旨在保留原始图像的结构和光照,同时加入流畅的镜头运动与逼真的转场,从而确保一致性和高质量结果。
可以。你可以提供 ‘last_image’ 作为转场或结束帧,从而增强生成视频的叙事连贯性。