Kling AI

Kling O1

Kling O1 is a unified multimodal video model for generating new shots and modifying existing footage. It can work from a prompt alone, animate first and last frames, incorporate image references, follow a video reference for style or camera movement, or treat that video as the base for an edit while preserving its timing and optional original sound.

How to use this model

  1. 1

    Choose one workflow: text generation, first-frame animation, first/last-frame interpolation, reference-guided generation, or base-video editing.

  2. 2

    Attach only the media needed for that workflow; an end image requires a start image, and editing requires a reference video set as the base.

  3. 3

    Identify every image and video reference in the prompt and state whether it contributes a subject, environment, style, action, or camera movement.

  4. 4

    Select Standard or Pro, then set aspect ratio and duration where the chosen inputs allow those controls.

  5. 5

    For edits, describe the requested change and explicitly name the motion, timing, framing, and sound that must remain unchanged.

O1Available

O1

This Kling O1 entry accepts a prompt plus optional start and end images, up to seven image references, and one selected 3–10 second reference-video segment. Feature mode uses the video for style or camera guidance; Base mode edits it directly, preserves its duration, and can retain its original sound.

Try now

Properties

  • Inputs: prompt, optional start/end images, up to seven image references, and one optional 3–10 second video segment.
  • Reference video supports Feature guidance or Base editing; Base mode can preserve original sound and inherits the selected clip duration.

How to use this model

  1. 1

    Use no media for text-to-video, add a start image for image-to-video, or add both start and end images for interpolation.

  2. 2

    Add up to seven reference images without a video, or up to four when a reference video is present.

  3. 3

    Set the video reference type to Feature for style or camera guidance, or Base to edit the selected source segment.

  4. 4

    Choose quality and set a 5- or 10-second duration without video; a reference video supports any whole-second duration from 3 to 10 seconds.

  5. 5

    When editing a base video, decide whether to keep its original sound and verify that the prompt preserves the intended motion and timing.

Parameters

Prompt

prompt
Required

Text prompt for video generation or editing. Can include references such as <<<image_1>>> and <<<video_1>>>.

Start Image

start_image
Optional

First frame image for the video. Supports .jpg, .jpeg, and .png files up to 10 MB.

End Image

end_image
Optional

Last frame image for the video. Can only be used when a start image is provided.

Reference Images

reference_images
Optional

Reference images for elements, scenes, or visual styles. Up to 7 images without a reference video, or up to 4 with one.

Reference Video

reference_video
Optional

Reference video for visual style, camera movement, or as the base for editing. The selected clip must be between 3 and 10 seconds.

Video Reference Type

video_reference_type
Optional

Choose whether the reference video guides the style and camera movement or serves as the base video to edit.

Default: feature

  • Style or camera reference

  • Video editing

Keep Original Sound

keep_original_sound
Optional

Preserve the original audio track from the reference video.

Default: true

Quality

mode
Optional

Standard mode is more cost-effective, while Pro mode provides higher output quality.

Default: pro

  • Standard

  • Pro

Aspect Ratio

aspect_ratio
Optional

Aspect ratio of the generated video. Ignored when using a start image or base video editing.

Default: 16:9

  • 16:9

  • 9:16

  • 1:1

Duration

duration
Optional

Output video duration in seconds. Ignored when editing a base video.

Default: 5Range: 3 – 10