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



AI model preview image
The DiffBIR model is designed for blind image restoration, employing a generative diffusion approach. The model takes an image as input along with a set of parameters such as seed, steps, tile size and stride, guidance scale and space, use of guidance, color fix type, etc. The model also allows the user to select the face detection model, upscaling model, and restoration model types. After processing, the model outputs a series of restored images in PNG format. The objective is to enhance the quality of the image through super-resolution, restoration, and upscaling, even when the original quality and details are unknown.

Use cases

The AI model 'DiffBIR' has implications for a wide variety of image restoration applications. With its capabilities in blind image restoration using generative diffusion prior, it has potential in domains such as digital photography, surveillance, and medical imaging where image quality is paramount. For example, professional photographers could use this technology to repair and enhance older or damaged photos. Furthermore, the model's ability to improve image quality could be utilized in security surveillance systems to enhance video feeds, potentially enabling a higher identification rate. Medical imaging could also benefit, as clearer and more detailed images can contribute to more accurate diagnoses. In terms of specific products, it could be incorporated into photo editing software, or used for developing standalone applications for image restoration and enhancement. Additionally, it's possible to integrate this AI capability into portable devices like smartphones or digital cameras, enabling users to edit and enhance photos on the go.


Cost per run
Avg run time

Creator Models

Animate Diff Scene Assembler$?52
Animate Diff$?28,356
Hello World$?39
Illuminati Diffusion$?3,539
Talknet Asd$?17

Similar Models

No similar models found

Try it!

You can use this area to play around with demo applications that incorporate the Diffbir 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 NameDiffbir
✨DiffBIR: Towards Blind Image Restoration with Generative Diffusion Prior
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-