Kling O3: Reference to Video
Upload a reference image and optional audio or text prompt. Generate a character-consistent video with native audio output and precise storyboard control.
概要
The 'Kling O3: Reference to Video' workflow is designed to transform a reference image into a character-consistent video, offering precise storyboard control and native audio output. This workflow leverages the Kling O3 model, renowned for its ability to maintain character consistency across frames while integrating optional audio or text prompts for enhanced narrative depth.
入力するもの
- ワークフローを読み込み中
ComfyUI で実行できます。 このワークフローはこのページでは実行できません。すべてが読み込まれた状態でComfy Cloudで開くか、ご自身のComfyUI用にダウンロードしてください。
Cloudで試すこのワークフローにはまだサンプル出力がありません。
よくある質問
Yes, by enabling the BatchImagesNode with Ctrl+B, you can process multiple reference images simultaneously.
No, adding an audio or text prompt is optional, but it can enhance the narrative and engagement of the final video.
The Kling O3 model is specifically designed to maintain visual consistency across frames, ensuring that the character's attributes remain stable throughout the video.
The LoadImage node supports common image formats such as JPEG, PNG, and BMP.