Get a weekly rundown of the latest AI models and research... subscribe!

Gfpgan Video


AI model preview image
The gfpgan-video model is designed for upscaling human face videos. Its primary function is to enhance the quality and resolution of human face videos, making them clearer and more detailed. The model takes a video file as input, with adjustable scale parameters for video upscaling. It uses GFPGAN (Generative Face Prior Guided Attention GAN) technology which is a deep learning model to perform these tasks. The output provided by the model is a URL to the upscaled video file.

Use cases

The GFPGAN video AI model is essentially designed for human face video upscaling. Essentially, it improves the quality of videos featuring human faces, making them clearer and more detailed. This can be applied to a variety of use cases, predominantly in sectors requiring high-quality video data. For instance, the entertainment and film industry might utilize the model to enhance old or poor-quality video clips. In surveillance and security, this could be used to upscale CCTV footage to help identify individuals more clearly. In the social media domain, the model could assist users in improving the quality of their uploaded videos. Regarding possible products, companies could offer video enhancement software or apps, where users upload their video and get an improved version as output. Another potential application could be in video conferencing platforms to upscale the video quality in real-time, providing a better experience for users. The model might also be integrated into video editing software, adding value by improving the final video quality. Given the advancement in deepfake technology, GFPGAN might be useful in improving the quality of faces in the generated videos. The model could also be incorporated into a service to restore and digitize old films or personal footage.


Cost per run
Avg run time

Creator Models

No other models by this creator

Similar Models

No similar models found

Try it!

You can use this area to play around with demo applications that incorporate the Gfpgan Video model. These demos are maintained and hosted externally by third-party creators. If you see an error, message me on Twitter.

Currently, there are no demos available for this model.


Summary of this model and related resources.

Model NameGfpgan Video
GFPGAN for human face video upscaling
Model LinkView on Replicate
API SpecView on Replicate
Github LinkView on Github
Paper LinkView on Arxiv


How popular is this model, by number of runs? How popular is the creator, by the sum of all their runs?

Model Rank
Creator Rank


How much does it cost to run this model? How long, on average, does it take to complete a run?

Cost per Run$-
Prediction Hardware-
Average Completion Time-