Face Swap places a face from one photo onto another image or a newly generated scene. Provide a face source and a target — either an existing photo or a text description — and get a photorealistic result in seconds.
It works in two modes: swap onto an existing image (the body, pose, and background are preserved, only the face changes) or generate a brand-new scene around the face using a text prompt. You can combine both — pass a target image and a prompt to guide how the swap is applied. Useful for marketing mockups, creative content, and visualizing how someone looks in different contexts.
What you can do
- swap_face — swap a face from a source photo onto a target image or a generated scene described by a prompt
Who it's for
Marketing teams creating personalized content, content creators experimenting with visual styles, and developers building photo editing or virtual try-on features.
How to use it
- Get a URL for the source photo containing the face you want to use.
- Call swap_face with face_image and either target_image (to swap onto an existing photo) or prompt (to generate a new scene).
- Optionally specify aspect_ratio and output_format. Use list_models to see available model options.
- Display the result inline with the returned image URL.
Getting started
Call swap_face with a face_image URL and a target_image URL to see the swap result immediately.
