This guide shows how to generate videos using LTX Video 2.3 models via the Leonardo.AI REST API.

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 '{
  "request": {
    "model": "ltxv-2.3-pro",
    "public": false,
    "parameters": {
      "prompt": "<Your_PROMPT>",
      "mode": "RESOLUTION_1080",
      "prompt_enhance": "AUTO",
      "quantity": 1,
      "duration": 8,
      "audio": true,
      "width": 1920,
      "height": 1080,
      "guidances": {
      "start_frame": [
        {
          "image": {
            "id": "<YOUR_START_IMAGE_ID>",
            "type": "GENERATED"
          }
        }
      ],
      "end_frame": [
        {
          "image": {
            "id": "<YOUR_END_IMAGE_ID>",
            "type": "UPLOADED"
      ]
    }
  }
}'

Recipe

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
model	`string`	Specifies the model used for generation. Set to `ltxv-2.3-pro` or `ltxv-2.3-fast`.
public	`boolean`	Controls whether the generated video is public (`true`) or private (`false`).
prompt	`string`	Text description of the video content to generate.
prompt_enhance	`string`	Controls whether prompt enhancement is applied. Set to `ON`, `OFF`, or `AUTO`.
quantity	`number`	Number of videos to generate per request. Max `1`.
duration	`number`	Length of the generated video in seconds. Pro: `6`, `8`, `10` (default: `8`). Fast: `6`, `8`, `10`, `12`, `14`, `16`, `18`, `20` (default: `8`).
seed	`number`	Optional. Seed value for reproducible generation.
width	`number`	Sets the width of the video in pixels. Default `1920`.
height	`number`	Sets the height of the video in pixels. Default `1080`. Set `width: 1080, height: 1920` for 9:16 portrait.
mode	`string`	Resolution tier. Set to `RESOLUTION_1080` (default), `RESOLUTION_1440`, or `RESOLUTION_2160`.
motion_has_audio	`boolean`	Whether to generate audio for the video. Default `true`.
guidances.start_frame	`array`	Optional. Single image to use as the starting frame (image-to-video). Max 1 item.
guidances.end_frame	`array`	Optional. Single image to use as the ending frame (transition video). Requires `start_frame` to also be set. Max 1 item.

Dimensions

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

Text-to-Video

Resolution	Aspect Ratio	Pixel Values (Width × Height)
1080p	16:9	1920 × 1080
1080p	9:16	1080 × 1920
1440p	16:9	2560 × 1440
1440p	9:16	1440 × 2560
2160p	16:9	3840 × 2160
2160p	9:16	2160 × 3840

Image-to-Video

Image-to-video may specify height and width for cropping, or set both to 0 to default to the source image's aspect ratio.

Note: When width and height are not specified, and guidances.start_frame is specified, the generation will base the dimensions on the guidances.start_frame image.