AI Video Generation

Start with an existing image

Open the intended gallery image and choose Generate Video. This is an image-to-video workflow, not a direct text-to-video form. Your written instruction describes the motion of the existing scene.

For a short clip, choose one action. An adult character turning toward a doorway is easier to judge than a sequence that changes location, outfit and camera position several times.

Choose duration and quality

Sound is a mode choice

The generator distinguishes Video + Sound from a silent mode. Choose a sound-enabled result deliberately instead of assuming that every generated clip has an audio track.

After a render, review both picture and sound. Motion quality, timing and the accuracy of a spoken result can vary. Music generation is a separate tool, not a promise that a video will contain a specific soundtrack.

Premium does not remove every video limit

Video generation requires Premium access. Deluxe Plus lists an included quota of 50 videos across quality levels; the plan comparison does not list an included video quota for Premium or Deluxe. Review the allowance and any Souls charge shown for the request rather than treating unlimited chat or images as unlimited video.

Rendering runs as a background job. Check its status before submitting another copy of the same request. The documentation says failed render jobs refund the Souls used; a result you dislike is not necessarily a failed job.