Put one face into any video clip. Upload the footage and the face you want in it - the original movement and timing are preserved frame by frame.
Face swap examples
One face swapped in, the original performance kept.
A video face swap puts one person's face onto the body and performance of someone in a video clip. The original movement, expressions and timing are kept frame by frame - only the face is replaced, which is why the result looks like the new person doing the old performance rather than a new video.
The clip you want the face to appear in. MP4, MOV or AVI.
A clear, well-lit photo works best; front-facing beats an angle.
Head movement and expressions from the original clip are kept.
Free runs are 720p with a watermark. Paid plans give clean exports.
New accounts get 30 credits, enough for five swaps. No card.
The face changes; the movement, timing and expressions do not.
Nothing has to come from a stock library.
Runs in the browser on any machine.
The most common use by a wide margin.
Swap your own face into a clip you like.
See how a different face reads in a shot you already have.
Our AI uses advanced deep learning algorithms to detect facial features in both the source and target images/videos. It then creates a realistic mapping of the source face onto the target, maintaining natural expressions, lighting, and movement while ensuring seamless integration.
For source faces, we support JPG, PNG, and WebP image formats up to 10MB. For target content, we accept both images (JPG, PNG, WebP) and videos (MP4, MOV) up to 20MB. All files should have clear, visible faces for best results.
Yes, there is a free tier. It covers short clips at lower resolution and adds a watermark. Paid plans raise the resolution and length, process faster, and remove the watermark — the pricing page has the details.
Face swap video generation typically takes 2-4 minutes, depending on the complexity of the content and current server load. Processing time may vary based on video length, resolution, and the number of faces detected in the target content.
The best source photos are high-resolution, well-lit images with the person facing forward or at a slight angle. Avoid heavily shadowed, blurry, or low-quality images. The face should be clearly visible without obstructions like sunglasses, masks, or hair covering facial features.
Currently, our AI focuses on the most prominent face in the target content for optimal quality. For multiple face swaps in group content, you may need to process each person separately. We're working on multi-face swap capabilities for future updates.
Your uploads are used only to produce the video you asked for. They are not sold, shared, or used to train anything. Our privacy policy is the authoritative statement on what we store and for how long.
Only use images of people who have agreed to it. Do not make anything misleading, harmful or defamatory, and do not use it on public figures to put words or actions on them they never said or did. Our terms of service cover what is and is not allowed.