Record and Download a Live Face Swap in Your Browser

Use your webcam and a reference image, generate with credits, record with optional microphone audio, replay and save your result.

Last updated: 2026-11-01

Live Face Swap at live-face-swap.net transforms a live webcam feed. You can record the generated result and download it in the same browser. A reference photo sets the appearance; an existing video cannot be uploaded as the source footage. No desktop application or local AI model installation is required.

Before you open the camera

Use a browser that permits camera access, live video connections and recording. Keep a stable internet connection and enough device storage for the file.
Sign in and check your available credits, selected AI engine and displayed rate. Camera preview is free; live AI generation uses credits, including time when you are not recording.
Prepare one reference image you own or have permission to use. JPG, PNG and WebP are accepted, up to 4 MB and 16 megapixels. Use a clear, evenly lit face. LFS Mirror requires an image at least 512 × 512 pixels; follow the upload hint for the selected engine.
Review the data-processing notice and likeness rules. During generation, the reference and camera video go to the selected model provider.

The AI runs through a cloud service. This does not establish a minimum hardware specification or guarantee that every laptop, phone or browser can generate and record reliably. The workbench checks recording support; a complete device compatibility matrix and measured latency or FPS results are not published.

1. Choose the reference and engine

Open the Face Swap studio. Choose a built-in look or upload your reference. Selecting a reference does not start generation. LFS Prism supports Face Swap; LFS Mirror supports Face Swap, Restyle and Try-On. See the engine rates and plan time estimates before starting.

2. Open the camera and start live AI

Choose Open camera and grant camera permission. Check lighting and framing. Read the displayed usage rules and rate before starting generation; use the Face Swap control when prompted. Restyle and Try-On ask for confirmation before generation. Wait for a moving generated result before recording. The original camera preview alone is not an AI result.

If sign-in, credits or the selected service are unavailable, follow the message shown in the workbench. Do not assume that turning on the camera has started generation: check the current session state.

3. Record a take, with or without your voice

Choose Start recording after generated video appears, or enable Auto-record to capture the live result when available. Earlier unrecorded footage cannot be recovered. Optional microphone audio is off by default; enable it deliberately and allow microphone access when asked. Your microphone is recorded locally and is not sent to the AI model. The visual effect does not change your voice.

Choose Stop recording to finish the clip. Stopping a recording does not end AI generation or its charges. Choose End session when you are finished generating. Ending a session also stops recording and closes the camera.

4. Replay, download and check the file

Wait for the clip to finish preparing, replay it and choose Download video. The recording contains the generated video, with microphone audio only if you enabled it. Check that the file has actually saved and plays on your device; a download request alone does not confirm a saved file.

The recorder tries MP4, then supported WebM formats, then the browser's default format. The file extension follows the format the browser actually produced. MP4 is not guaranteed, and this page does not offer a separate transcoding or trimming tool. If the download action does not save the file, open the video and use your browser's save or share action.

Save before refreshing, leaving or closing the tab. Recordings and replay URLs are temporary in this browser tab; the site does not upload recordings or keep a recovery copy. Save a take before recording another one. A large recording can reach the browser recorder's size limit; save shorter takes separately.

Understand what uses credits

Reference selection, camera preview, recording, replay and download add no AI charge. LFS Prism bills from an established realtime connection until it disconnects, including connected time before the first generated frame. LFS Mirror bills generated seconds, rounded up. Ending a session starts usage reconciliation; unused reserved credits return after settlement. See current rates and reservations and the billing terms.

Troubleshooting

No camera: check site permission, confirm that a camera is connected, and close other software using it.
Camera visible, no generated result: inspect the workbench's sign-in, credits, connection and error status. Recording requires a live AI output.
Recording unavailable: the browser may lack the recording API or a usable encoder. A compatible screen layout does not guarantee recording support.
No microphone audio: enable audio before recording and check microphone permission. Playback of an existing silent recording cannot recover missing audio.
Download missing: keep the tab open, retry saving, or open the video and use the browser's save action. Recordings cannot be recovered after leaving.
Usage pending after stopping: return to the workbench or credit history to retry reconciliation; contact support with the session identifier if it remains pending.

Can this replace my camera in OBS, Zoom or Discord?

This site provides browser creation and local recording. It does not provide a virtual camera or a verified output path into those apps.

For help, email support@live-face-swap.net with your device, browser and error message. Do not attach private face images, payment-card details or identity documents.