Skip to main content

How do I use sketch to show invideo agent the composition I want?

Draw a visual reference in agent chat and use it to guide image or video generation.

Type /sketch in the AI agent chat to open the drawing space. Draw freehand, add shapes, or upload a photo and annotate it. You can layer these elements to show where subjects, objects and text should sit.

Attach the finished sketch to your message like an image, then explain what the agent should take from it: framing, placement or composition.

The agent can use the sketch when generating images and videos. For example, use shapes to show the subject on one side and space for a title on the other. Review the result and refine your direction as needed.

Did this answer your question?