MobileFaceSwap generates a new face using your photo’s identity and the target video’s motion and expression. Assign each person once to apply it to detected appearances. Files stay on your device. Up to 30 seconds, a 720-pixel longest edge and 15 fps output. No clothing changes. The editor is available in Turkish and English.
Comment utiliser AI Video Face Swap (Beta)
- Choose a video file or an accessible direct video URL and select up to 30 seconds.
- Faces are detected automatically. Add one photo per person, check the matches and merge cards if needed.
- Create the video, check the result and audio, then download the MP4 or WebM file.
Questions fréquentes
Is it free and are my files uploaded?
Free for noncommercial experiments. Face detection, neural face generation and video encoding run on your device. Videos and photos are not uploaded. About 360 MB of models and runtime files download on first use.
Does it work on every phone?
The browser must support WebCodecs and the model runtime. Current Chrome or Edge is recommended. Speed, memory and heat depend on the phone. Unsupported devices cannot start processing.
Does it generate realistic faces or change clothing?
MobileFaceSwap generates a 224-pixel face with the target video’s motion and expression. Profiles, occlusions, fast movement and missed detections can leave incomplete results. Clothing and hair changes are not supported. No commercial-use license is granted.