Swap a face in a video

Video Face Swap transfers a single photo's face onto every frame of a target video. It keeps the source hairstyle, head shape, and proportions while adopting the target's expressions and lighting, so the face sits naturally in the scene. Identity stays locked from the first frame to the last — no drifting, no flicker, no mask floating on top.

Open Video Face Swap

Open Apps from the sidebar and click Video Face Swap. The tool opens with two input slots on the right — Step 1 for the source face, Step 2 for the video you want to map it onto.

Add the source face and target video

Under Step 1: Source Image, upload a single clear, front-facing photo of the face you want to use. Under Step 2: Target Video, add the video you want the face mapped onto — upload it or pick it from your library. A flat, evenly lit portrait and steady target framing give the cleanest transfer.

Video Face Swap interface — a source face in Step 1 and a target video in Step 2

One source photo is enough — no training and no dozens of angles. The model maps that single face across the whole video.

Generate and download the video

Click Generate. The face is transferred frame by frame, tracking motion and expression so the identity holds throughout. Preview the finished video, then download it.

Generated result in Video Face Swap with the source face mapped onto the target video

Each generation costs 7 credits. The swapped face adopts the target's mouth movement and lighting, so it stays consistent as the subject turns or speaks.

To swap a face in a single image instead, use Image Face Swap. Sharpen the output with Video Upscaler, or learn the basics of video in Generate a video and Video result actions.