Free · Runs on your device

Depth Map Video Generator

Turn a short video into a grayscale depth video. Processing stays in your browser and uses your device GPU when available.

Drop a video here or click to choose

MP4, WebM or MOV · up to 50 MB · first 10 seconds

Output settings

ResolutionAnalysis FPSOutput formatMP4 is the most compatible choice. WebM usually encodes faster.Invert depthSwitch which side of the depth range appears white.

The first run downloads an approximately 50 MB model and caches it in your browser.

Depth Anything V2 Small · Apache-2.0

Drop a video here or click to choose

Private by design

Your source video is not uploaded to Z-Image. Model inference and video encoding happen on this device.

Stable short-video output

Depth Anything V2 Small estimates each frame, while temporal smoothing reduces brightness flicker and resets automatically at scene cuts.

Browser compatibility

Chrome or Edge with WebGPU is fastest. Other modern browsers use a slower WebAssembly fallback when supported.

Why use a depth map video?

A depth sequence separates scene geometry from color and texture. It gives AI video workflows a cleaner signal for relative distance, occlusion and motion while prompts or reference images control appearance.

Preserve spatial structure

Keep the foreground, background and object boundaries easier to distinguish when rerendering or changing style.

Guide motion over time

Consistent depth frames can help downstream models follow scene layout, parallax and camera movement with less drift.

Reuse one geometry pass

Use the exported sequence as a depth condition, mask or starting point for depth of field, fog, compositing and new-view workflows.

How to create a depth map video

Upload a short video

Choose an MP4, WebM or MOV file. Only the first 10 seconds are processed.

Choose speed and detail

360p and 8 FPS are faster; 480p and 12 FPS preserve more motion detail.

Process and download

Keep this tab open while the browser estimates, smooths and encodes the depth frames.

Depth map video questions

What is a depth map video?

It is a sequence in which pixel brightness represents relative distance from the camera. A common convention is brighter for nearer areas and darker for farther areas, and this tool lets you invert that range.

How is a depth video used in AI video generation?

Compatible workflows can use it as a geometry condition alongside a text prompt or reference image. The depth sequence guides layout and motion, while the other inputs define subject appearance and style.

What can I do with the exported depth sequence?

Typical uses include video rerendering, style transfer, ControlNet-style depth guidance, masks, depth of field, fog, compositing and experimental new-view generation.

Is this tool free?

Yes. The free mode runs locally on your device and does not use Z-Image credits.

Why is the first conversion slower?

The browser downloads and prepares the depth model once. Later runs can reuse the cached files.

Is this metric depth?

No. This release produces relative depth for visual effects and masks, not real-world distance measurements.

Does the output keep audio?

No. A depth map is an analysis pass, so the exported video is silent.