Create video with image sora-2

POST

/v1/video/create

Use the sora-2-all video model to create a unified-format video from multiple images and a prompt.
Suitable for quickly turning static images into portrait or landscape short videos.

Authorizations

bearer

Type

HTTP (bearer)

Request Body

application/json

object

Array of image URLs. Each element must be a publicly accessible image URL.

Model name, for example 'sora-2-all'.

Video orientation:

• portrait: vertical
• landscape: horizontal

Prompt describing the content and style of the video.

Video size, e.g. 'small' (around 720p), 'large', etc.

Video duration in seconds, e.g. 10 or 15.

Watermark control:

• Default true: prefers a watermark-free video; falls back to a watermarked version if removal fails.
• Pass false: force watermark-free video; will automatically retry when watermark removal errors occur.

Responses

Video creation task created successfully

Content-Type

application/json

object

Task ID used to query task status and result.

Task status, such as 'pending', 'processing', 'succeeded', or 'failed'.

Timestamp in milliseconds when the status was last updated.