Face swap video prompt and editing guide

What you are making

The finished film follows a short clip you upload. One adult portrait replaces the face of the person in that clip.

The planned result keeps the original body, clothing, motion, camera, and background. Models do not always hit that exactly. Judge the result by whether the face stays recognizable and the clip is still the same action.

Files that work

Use one portrait of one adult. Keep the eyes, hairline, and expression visible, with even light and little covering the face.

Use an MP4 or MOV clip of that same kind of shot: one adult, face visible, 4–15 seconds. A group photo or a clip with several people gives the model extra faces to chase.

After you download

Watch the face first. Is it still the person in the photo, and does the body still follow the original clip?

Then listen. The file has new scene audio and does not keep the original soundtrack, so add music or captions in your editor if the post needs them.

Placeholders {{duration}}, {{aspect_ratio}}, and {{resolution}} are filled when a video is purchased. @Video1 is the original video. @Image1 is the replacement face.

@Image1 is one adult face. @Video1 is the original clip. Replace the face of the single adult in @Video1 with the face in @Image1. Keep that person's body, clothing, motion, camera, and background from @Video1. Do not add a second person. Do not depict a minor. Do not invent a new scene.

The film is {{duration}} seconds at {{aspect_ratio}}, {{resolution}}. The new face stays matched to @Image1 for the whole clip: eyes, hairline, and face shape. No sunglasses, masks, or hair covering the eyes. No captions, logos, or on-screen text. Sound is scene audio that fits the action already in @Video1. Do not copy a song from @Video1. No added music.

Check the result

• The face matches the uploaded photo.
• The body and motion still follow the original clip.
• Only one person is in frame.
• The background stays with the original clip.
• Scene audio does not jump harshly.