Sample Request

curl --request POST \
     --url https://cloud.leonardo.ai/api/rest/v2/generations \
     --header 'accept: application/json' \
     --header 'authorization: Bearer <YOUR_API_KEY>' \
     --header 'content-type: application/json' \
     --data '
{
    "model": "wan-2.6",
    "public": false,
    "parameters": {
      "prompt": "The koala plays with the cat",
      "guidances": {
        "start_frame": [
          {
            "image": {
              "id": "<YOUR_START_IMAGE_ID>",
              "type": "UPLOADED"
            }
          }
        ]
      },
      "duration": 5,
      "negative_prompt": "",
      "width": 1920,
      "height": 1080
    }
}
'

API Request Endpoint, Headers, Parameters

Endpoint

https://cloud.leonardo.ai/api/rest/v2/generations

Headers

--header "accept: application/json" \
--header "authorization: Bearer <YOUR_API_KEY>" \
--header "content-type: application/json"

Body Parameters

Parameter	Type	Definition
duration	`number`	Duration of the video in seconds. Set to `5`, `10`, or `15`. Defaults to `5`.
guidances.audio_reference	`array`	Maximum 1 item. Each item includes audio.id and audio.type (`UPLOADED`). Specifies an audio reference for guidance.
guidances.start_frame	`array`	Maximum 1 item. Each item includes image.id and image.type (`INIT`, `GENERATION`, `UPLOADED`, `GENERATED`, or `VARIATION`).
guidances.video_reference_base	`array`	Maximum 3 items. Each item includes video.id and video.type (`UPLOADED` or `GENERATED`). Base video to be edited.
height	`number`	Sets the height of the output video in pixels. See Dimensions for valid width and height combinations.
model	`string`	Specifies the model used for generation. Set to `wan-2.6`.
negative_prompt	`string`	Text description of content to avoid in the generated video. Maximum of 500 characters. Defaults to `""`.
prompt	`string`	Text description of the video content to generate. Maximum of 1500 characters.
public	`boolean`	Controls whether the generated video is public (true) or private (false).
quantity	`number`	The number of outputs to generate. Set to `1`. Defaults to `1`.
resolution	`string`	Output resolution when using AUTO dimensions (`width: 0`, `height: 0`) for image-to-video. Valid values are `720p` and `1080p`. Determines cost tier. Defaults to `1080p`.
seed	`number`	Random seed for reproducible generation. Set between `0` and `2147483647`.
width	`number`	Sets the width of the output video in pixels. See Dimensions for valid width and height combinations.

Dimensions

The API accepts width and height parameters in the following aspect ratio combinations:

Resolution	Aspect Ratio	Pixel Values (Width x Height)
720p	16:9 (Landscape)	1280x720
	4:3	960x720
	1:1 (Square)	960x960
	3:4	720x960
	9:16 (Portrait)	720x1280
1080p	16:9 (Landscape)	1920x1080
	4:3	1440x1080
	1:1 (Square)	1440x1440
	3:4	1080x1440
	9:16 (Portrait)	1080x1920
Auto		0x0

Setting width: 0 and height: 0 (AUTO) lets the model infer dimensions; use the resolution parameter to control the output resolution tier.