Video Generation With Audio

Kling 3.0 Pro Image-to-Video

Generate cinematic videos from images using fal-ai/kling-video/v3/pro/image-to-video with custom element support.

Typical generation time: 60-180 seconds

Model

Start Image

  • Uploaded image

Upload or paste URL of the starting image

End Image (Optional)

Maximum file size: 10 MB

Upload or paste URL for the end frame

Video duration in seconds (3-15).

How closely to follow the prompt (0-1)

Characters/objects to keep consistent in the video. Each element is an image set (frontal + reference images) or a video; it may also include a voice_id (from the Kling create-voice endpoint) — the voice is bound to that element, and prompt references to the element speak with it. Reference in prompt as @Element1, @Element2.

Multi-shot storyboard (1-6 shots), each with its own prompt and duration (total ≤ 15s). Mutually exclusive with Prompt.

How multi-shot structure is determined. 'Intelligent' lets the model decide the shot breakdown.

Explore example prompts and their results.

  • Example 1Video
  • Example 2Video
  • Example 3Video

Similar Models

Community