Read the visual language inside any video.

Upload a short clip and turn its subject, composition, camera, light, color, and style into model-ready prompts.