Cloud Software
Deepfakes App is online deepfake software that works in the cloud. All you need to do is upload videos and click a button, our app does the rest.
Private
Only you have access to your learning data, videos and images. We do not share information with any 3rd parties or publish your videos in any way.
Reuse model
You can reuse your trained model. This allows you to improve the face swapping quality of the results, or make additional videos without re-training a model..
* The quality of your deepfake depends on the quality and length of your videos. The more well trained your AI model is, the better the result.
* Please use our technology responsibly and review Our content policy and our terms of use.
How to Make a Deepfake Video
Step 1. Upload Your Source & Target Videos
Step 2. Let the AI Learn and Render your Video
3. Download or Watch your Newly Created Deepfake Video
You can now watch the video.
4. Reuse your model
The lower the loss values are, the higher the quality will be. (They should be less than 0.02)
Pricing
$3 / hour
Average Cost and Time
Basic Deepfake (10,000 Iterations) – $15
High-Quality Deepfake (50,000 Iterations) – $60
* The quality of your deepfake depends on the quality and length of your videos. The more well trained your AI model is, the better the result.
* Please use our technology responsibly and review Our content policy and our terms of use.
Responsible Deepfake Technology
Visible Watermarks
Every deepfake generated on our tool has a clear and visible watermark indicated that the video is a deepfake. We also leave clear traces of manipulation in the video data so it’s easy to identify as fake. We believe deepfake technology should be clearly labelled.
Imperfect by Design
Deepfake technology is incredibly advanced and can easily fool humans. We intentionally don’t push the limits, so deepfake videos can be enjoyed while still being able to identify it’s fake through the imperfections.
Reuse model
You can reuse your trained model. This allows you to improve the face swapping quality of the results, or make additional videos without re-training a model..