Upload a video and get a depth map video that stays stable from the first frame to the last — no flicker, no brightness pumping. Use it to recreate the camera moves and motion of any clip with depth-controlled video generation. Free credits on signup.
It is a grayscale version of your video in which brightness encodes distance: white is closest to the camera and black is farthest. It captures the 3D layout of the scene and how the camera and subjects move through it.
The most popular use is recreating the camera moves and motion of a reference clip: feed the depth video to a depth-controlled video model and generate new content that follows the same movement. It is also used for depth-of-field, fog, relighting, compositing and AR occlusion.
No. Depth is estimated across groups of frames together and normalized once over the entire video, so a subject keeps the same gray level as the camera moves and the scene changes.
Grayscale is the standard depth map and the one to use for generation and compositing workflows. Inferno and Turbo are color visualizations that make depth differences easier to see.
Up to 10 minutes per video. The depth video keeps the input resolution and duration; frame rates above 30 fps are returned at 30 fps.
$0.005 per second of video, with a 3-second minimum. The exact cost is shown before you generate, and new accounts get free credits.