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

Sdxl 2004



The sdxl-2004 is a model specifically designed for image-to-image tasks. Developed based on digital photography from 2004, it can transform input images according to specific prompts. The model has been fine-tuned to accept an input of image parameters such as width, height, prompt, refining parameters, guidance scale, noise level, negative prompt, and output strength. It produces an image as an output corresponding to these input parameters. The objective is to enhance, refine, and stylize the image in a specific way, as indicated by the input prompts, all while maintaining a 2004-styled aesthetic. The model doesn't apply any watermarks on the resulting outputs.

Use cases

The sdxl-2004 is an Image-to-Image AI model whose strength lies in refining low-quality images, specifically those produced by older digital photography such as from 2004. It appears this tool could be particularly useful for various fields, including but not limited to digital archiving, restoration, and preservation, where there is the need to enhance the visual quality of older, degraded images. This AI could allow for the recovery and improvement of antique digital photographs, bringing them up to modern visual standards. Additionally, industries like filmmaking or documentaries, where there's a need to incorporate older digital photography, could find this model quite useful. It might enable them to incorporate historic or older footage with newer, sharper imagery without a noticeable quality gap. In terms of practical applications, software or applications focused on digital image restoration could leverage this model. It would allow users to upload old or degraded digital images and receive an improved output. Another potential application could be in the creation of unique visual arts or decorations - giving a "retro-modern" look to the imagery, akin to the cyberpunk style that the model appears to handle well. Lastly, in computer vision and AI research, this model could serve as a tool to better understand the progression of digital photography and image quality over time.


Cost per run
Avg run time

Creator Models

Sdxl Santa Hat$?475
Star Trek Adventure$0.0483154
Sdxl Dots$?167
Sdxl Cats Movie$?479
Sdxl Barbietron$?1,285

Similar Models

No similar models found

Try it!

You can use this area to play around with demo applications that incorporate the Sdxl 2004 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 NameSdxl 2004
An SDXL fine-tune based on bad 2004 digital photography
Model LinkView on Replicate
API SpecView on Replicate
Github LinkNo Github link provided
Paper LinkNo paper link provided


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-