How to Use AI Face Swap

Swap a face onto a photo or clip: upload the target, pick a source face, generate. Credit rates, how batch mode differs, and the consent and disclosure rules.

Face swap puts one face onto an existing photo or clip. Lighting and expression are meant to follow the target footage. It is not a model swap: the body stays, the face changes. Consent is the first input, not a later checkbox.

How to run a swap

  1. Upload the target in AI face swap: the image or video whose face you want to replace.
  2. Pick the source face. A saved AI character, or a reference photo you have rights to use. Front-facing, even light, the whole face visible.
  3. Generate. A still swap is 6 credits. Video is 10 credits per second. Review edges, teeth and hairline before you publish.

Many files at once

The same face across a set of files belongs in batch face swap (up to 50 files). The legal piece is the same: face swap versus deepfake.

Consent and labelling

Only swap a real person with that person's permission. Do not put a customer or a celebrity onto footage you do not have rights to. EU AI Act Article 50 requires you to disclose artificially generated or manipulated image, audio or video that constitutes a deep fake. Treat labelling as the default.

Frequently asked questions

Can I swap faces in video?

Yes. Stills are 6 credits. Video is 10 credits per second. Review the whole clip, not just the first frame.

How is this different from model swap?

Face swap replaces a face and leaves the body. Model swap rebuilds the person under a garment so the whole figure can change.

Do I need consent?

Yes, for any real person’s face. A saved AI character you created does not need a third-party likeness release. A customer or public figure does.

When should I use batch mode?

When the same source face has to land on many files. Batch takes up to 50 images or videos in one run.

Explore related tools