generateVideo(payload)

This method allows you to initiate a video generation task. It is available on both the ImagineoAIClient (for browser environments) and ServerImagineoAIClient (for Node.js server environments).

The method takes a payload object and returns a Promise that resolves with the video generation response, including the run_id for tracking the task.

Parameters

• payload (GenerateVideoPayload): An object containing the parameters for video generation. This object must conform to the GenerateVideoPayloadSchema.

  Key fields in payload:

  • prompt (string, optional): A textual description of the desired video content.
  • negative_prompt (string, optional): A textual description of what to avoid in the video.
  • image_url (string, optional): URL of an initial image to base the video on. Either image_url or source_run_id must be provided, but not both.
  • source_run_id (string, optional): The ID of a previous run whose output image will be used as the initial image. Either image_url or source_run_id must be provided, but not both.
  • aspect_ratio (string, optional): The desired aspect ratio of the video (e.g., "16:9", "1:1", "9:16"). Defaults to "16:9".
  • duration (number, optional): The desired duration of the video in seconds. Defaults to 3 seconds.
  • seed (number, optional): A seed for the random number generator to ensure reproducibility.
  • cfg_scale (number, optional): Classifier-Free Guidance scale. Controls how much the video generation should follow the prompt. Higher values mean stricter adherence.
  • motion_bucket_id (number, optional): Controls the amount of motion in the generated video.
  • fps (number, optional): Frames per second for the generated video.
  • webhook_url (string, optional): A URL to which a webhook will be sent upon completion or failure of the video generation.
  • webhook_secret (string, optional): A secret token to verify the authenticity of the webhook.

  Refer to GenerateVideoPayloadSchema in types.ts for the complete and most up-to-date schema definition.

Returns

• Promise<GenerateVideoResponse>: A Promise that resolves to an object containing details about the initiated video generation run.

  Key fields in GenerateVideoResponse:

  • run_id (string): The unique identifier for this video generation run.
  • status (string): The initial status of the run (e.g., "queued", "processing").
  • input (object, optional): The input parameters that were used for the generation.
  • error (object, optional): An error object if the request failed immediately (e.g., due to validation issues).

  Refer to GenerateVideoResponseSchema in types.ts for the complete schema.

Example Usage

import { ImagineoAIClient } from "@imagineoai/javascript-sdk";
// Or for server-side:
// import { ServerImagineoAIClient } from "@imagineoai/javascript-sdk/server";

const client = new ImagineoAIClient({
  auth: { apiKey: "YOUR_API_KEY" },
});

async function createVideo() {
  try {
    const videoPayload = {
      prompt: "A futuristic cityscape with flying cars",
      aspect_ratio: "16:9",
      duration: 5, // 5 seconds
      // You can also use an image_url or source_run_id
      // image_url: "https://example.com/your-image.png"
    };

    const response = await client.generateVideo(videoPayload);
    console.log("Video generation started:", response);
    console.log("Run ID:", response.run_id);
    // You can use the run_id to track the status of the video generation
    // using the getRunStatus method (if available and applicable).

  } catch (error) {
    console.error("Error generating video:", error);
  }
}

createVideo();

Notes

• Ensure that either image_url or source_run_id is provided in the payload, but not both. If neither or both are provided, a validation error will occur.
• The actual video generation is an asynchronous process. This method initiates the task and returns a run_id for tracking. You might need to poll an endpoint using this run_id or use webhooks to determine when the video is ready.