Create a video generation from an image

post

https://cloud.leonardo.ai/api/rest/v1/generations-image-to-video

This endpoint will generate a video using an uploaded or generated image.

Recent Requests

Time	Status	User Agent
Retrieving recent requests…

Loading…

Body Params

Query parameters can also be provided in the request body as a JSON object

prompt

string

required

The prompt used to generate video

imageId

string

required

The ID of the image, supports generated images and init images. Use only image or imageId with imageType.

imageType

string

enum

required

Type indicating whether the init image is uploaded or generated. Use only image or imageId with imageType.

Allowed:

endFrameImage

object | null

The end frame image. Supported only on kling2_1 model when used in conjunction with an init image.

resolution

enum

The resolution of the output video. Acceptable values vary based on model

Allowed:

model

enum

Defaults to MOTION2

The model to use for the video generation. Defaults to MOTION2 if not specified.

Allowed:

frameInterpolation

boolean | null

Smoothly blend frames for fluid video transitions using Interpolation.

isPublic

boolean | null

Whether the generation is public or not

seed

integer | null

Apply a fixed seed to maintain consistency across generation sets. The maximum seed value is 2147483637 for Motion 2.0.

negativePrompt

string | null

The negative prompt used for the video generation.

promptEnhance

boolean | null

Whether to enhance the prompt.

duration

integer | null

Duration of the output video in seconds. Allowed values: 4, 6, or 8 on models VEO3 and VEO3FAST (default: 8) and 5 or 10 on model KLING2_5 (default: 5).

promptEnhanceInstruction

string | null

A natural language instruction used to modify the main prompt. For example, 'make it cinematic', 'add a rainbow', or 'change the subject to a cat'.

styleIds

array of strings | null

Predefined styles to enhance the prompt. This accepts a list of style uuids.

Array of Strings

elements

array of objects | null

An array of elements/loras objects that will be applied sequentially to the output. Elements are only supported for Motion2.0 generations.

Array of Element Inputs

height

integer | null

Height of the output video. Acceptable values vary based on model

width

integer | null

Width of the output video. Acceptable values vary based on model

Response

200Responses for POST /generations-image-to-video