Deep-Live-Cam
Real-time face swap and video deepfake generation using a single image.
Perform real-time video face swapping and webcam deepfakes from a single source image.
Deep-Live-Cam is an open-source real-time face-swapping and video deepfake tool that operates from a single reference image. It suits developers, digital artists, and media creators who need rapid face replacement in live webcams or video files.
What it is
The software captures webcam feeds or processes video files to swap faces dynamically using ONNX models like inswapper and GFPGAN. It features specialized controls like Mouth Masking to preserve natural mouth movement and Face Mapping to handle multiple subjects simultaneously.
Who it's for
It suits AI media artists and developers experimenting with generative video pipelines. It does not suit beginners without technical skills unless they opt for the pre-built Ultimate version.
Key features
Real-Time Face Swap
Perform live video deepfakes and webcam swaps with a single reference image.
Mouth Masking
Retain your original mouth region for accurate movement during processing.
Multi-Subject Mapping
Apply different faces to multiple subjects simultaneously within the same frame.
Hardware Acceleration
Run efficiently via CUDA, CoreML, DirectML, or OpenVINO execution providers.
Up and running in 3 steps
Our honest verdict
What's great
- Operates in real-time on live webcam streams and pre-recorded video files
- Requires only a single reference image to generate the face swap
- Supports multiple hardware acceleration backends including NVIDIA, Apple Silicon, and Intel
- Includes built-in content checks to block inappropriate media processing
Where it falls short
- Manual installation requires deep technical skills, specific Python versions, and dependency management
- Requires external ONNX model downloads and proper configuration of execution providers
- The free version requires local hardware resources, while the Ultimate version requires purchasing via an external site
Pricing & license
The repository code is free under its open-source license, but the creators offer a paid pre-built Ultimate version with exclusive features and priority support on their official website.
Open remote AI-training roles
870 open remote AI-training roles, updated November 14, 2026; most pay $20–$170 an hour. Zekai lists them free; the application goes to the employer’s platform.
Alternatives
Use it, or have us set it up
FAQ
What hardware do I need to run this tool?
You can run it on a CPU, but GPU acceleration via NVIDIA CUDA, Apple Silicon CoreML, DirectML, or OpenVINO is strongly recommended for real-time performance.
Can I process live webcam feeds?
Yes, you can select your camera feed directly in the interface and press live to execute real-time face swapping.
