Skip to main content
Video generation is useful for marketing assets, product demos, short-form drafts, motion covers, and creative previews. Compared with text or image generation, video generation is often asynchronous: submit a task, receive a task ID, then query the result.

Workflow

1

Prepare a prompt or reference image

Describe the subject, camera movement, action, style, duration, and aspect ratio.
2

Submit a generation task

Use a supported task endpoint or dedicated video model.
3

Fetch the result

Poll by task ID and download or display the output when it completes.
Video generation is often billed per task or per output. Retries, resolution, and duration can affect cost.

Next steps