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



AI model preview image
The lama model is an image-to-image model that is used for large mask inpainting with Fourier Convolutions. It is designed to be resolution-robust and provides a method for imaging data gaps or missing information in images. The model takes an input of a masked image URL and a direct image URL, and through its Fourier convolution process, it delivers a URL of the restored image as output, filling the gaps with generated spatial details. Useful in areas like art restoration or image editing, this model can handle images with large missing parts and reconstructs them while maintaining resolution quality.

Use cases

The LaMa AI model, utilizing Fourier Convolutions, is designed to perform resolution-robust large mask inpainting. Essentially, it helps restore and complete parts of an image that might be obscured or missing. This image-to-image model could be exploited in an array of use-cases. In digital art and design, it could be a vital tool to auto-fill or correct images where the complete data isn't available. Similarly, for photo-restoration services, the model could assist in completing parts of old or damaged photographs. In surveillance system functionality, it could be applied to restore obscured images. The model can be leveraged in developing various software applications focusing on digital image correction, restoration, and processing. It may also be used in developing powerful tools for archeologists or historians, aiding in visual reconstruction of partially damaged ancient artifacts or documents.


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 Lama 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 NameLama
🦙 LaMa: Resolution-robust Large Mask Inpainting with Fourier Convolutions
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-