You set the style. deepshorts researches the topic, makes the assets, and assembles the piece.
You approve what ships.
Analysis, planning, generation, and rendering stay visible together. The edges carry the work from node to node. The rendered cut opens for an edit.
The line they remember.
EditYou do not start from a blank prompt. A format already knows how that kind of piece is made. You bring the topic and the style.
When the piece is assembled, it opens as a real timeline. Change the cut, the captions, the layout. The AI outcome is not locked in an export.
These four are the focus. The library is already 15+, and new formats are added continuously. A piece can be images, video, or both.
A person to camera, cut to the point. The performance stays. Video is the piece.
A sequence of images that teaches one idea. Add video when the explanation needs motion.
A story told in pictures, beat by beat. Any beat can be a video.
Motion, image, and type as one piece. Video is a native layer, not an attachment.