Model reference · open weights

swin2SR-realworld-sr-x4-64-bsrgan-psnr

Image caidas Image edit 1 build Open weights 13k dl/mo

swin2SR-realworld-sr-x4-64-bsrgan-psnr is an open-weight image model from caidas. swin2SR-realworld-sr-x4-64-bsrgan-psnr (BF16) weighs 48 MB; the smallest configuration that runs it is RTX 3060 12 GB.

  • Swin2SR-realworld-sr-x4-64-bsrgan-psnr is an image-to-image model developed by caidas for real-world image super resolution.
  • It upscales images by a factor of four and is based on the Swin2SR architecture introduced in the paper by Conde et al.
  • The model is distributed under the Apache 2.0 license.

Summary of the caidas/swin2SR-realworld-sr-x4-64-bsrgan-psnr model card, 2026-10-01

What it is

Released bycaidas
Released2022-12-16
VRAM48 MB for the weights

What it runs on

Memory and cards for swin2SR-realworld-sr-x4-64-bsrgan-psnr (BF16)

48 MBweights, file size
5.0 GBworking memory, one 1024² image
537 MBruntime overhead
CardOne imageMemory
RTX 3060 12 GBfits11.6 GB
RTX 4060 Ti 16 GBfits15.4 GB
RTX 3090 24 GBfits23.4 GB
RTX 4090 24 GBfits23.4 GB
RTX 5090 32 GBfits31.0 GB
L40S 48 GBfits44.0 GB
A100 80 GBfits78.2 GB
H100 80 GBfits78.1 GB
RTX PRO 6000 Blackwell 96 GBfits93.8 GB
DGX Spark (GB10) 128 GB unifiedfits107 GB
H200 141 GBfits138 GB
B200 180 GBfits176 GB

From the model card

What caidas says about swin2SR-realworld-sr-x4-64-bsrgan-psnr

Read the model card

Swin2SR model that upscales images x4. It was introduced in the paper Swin2SR: SwinV2 Transformer for Compressed Image Super-Resolution and Restoration by Conde et al. and first released in this repository.

Intended use cases

This model is intended for real-world image super resolution.

Usage

Refer to the documentation.

Quoted from the model card on Hugging Face. The full card is behind the Hugging Face link above.

Running it yourself

Run it on a rented GPU

Rent a machine by the hour. How to run this model is on its model card.

© 2026 AxForge · EU-hosted AI infrastructure Pricing Docs Trust Privacy Terms