Generate videos using the Hailuo 2.3 models via the Leonardo.AI REST API.

Sample Request

The following examples demonstrate how to generate videos using Hailuo 2.3 via the V2 generations endpoint.

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": "hailuo-2_3",
  "public": false,
  "parameters": {
    "prompt": "A cinematic shot of rain falling on a quiet city street at night",
    "mode": "RESOLUTION_1080",
    "duration": 6,
    "width": 1920,
    "height": 1080
  }
}
'

Recipe

API Request Endpoint, Headers, Parameters

This section defines the required endpoint, headers, and request structure used for all Hailuo 2.3 video generation requests.

Endpoint

All video generation requests are sent to the following endpoint.

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

Headers

Include the following headers in all requests to authenticate and define the request format.

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

Body Parameters

The request body defines the configuration of your video generation, including model selection, prompt, duration, resolution, and optional parameters.

Duration and Resolution

Duration and resolution are interdependent. Certain combinations are restricted based on model capabilities.

Aspect Ratio and Dimension

Aspect ratio behavior depends on whether you are using text-to-video or image-to-video.

Text-to-Video

Text-to-video outputs are fixed to a 16:9 aspect ratio.

Example: 1920 × 1080

Image-to-Video

The output video follows the aspect ratio of the input image used as the start frame.

No separate aspect ratio parameter is required

Dimension Reference

The following are commonly used width and height combinations based on supported aspect ratios.

Note: AUTO uses the dimensions of the input image (start_frame)

Styles

Use style_ids to apply predefined visual styles to your generated video.

Vibe

Lighting

Color Theme

Sample Request Using Style

You can combine styles by passing multiple style_ids, typically selecting one per category (Vibe, Lighting, Color Theme).

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": "hailuo-2_3",
  "public": false,
  "parameters": {
    "prompt": "A cat",
    "mode": "RESOLUTION_768",
    "duration": 6,
    "style_ids": [
       "28a53b2e-eb80-41d7-8360-208068fecf49",
       "3f705252-1197-4f59-b6ed-21625dce6a65",
       "f84bbf31-1d72-415d-ae4c-deb4b18fccba"
       ],
    "width": 1376,
    "height": 768
  }
}'