0
Image editing & upscalingBSD-3-ClausePython

Real-ESRGAN

Practical image and video super-resolution models.

GitHub
Stars
37.1K
Forks
4.5K
Open issues
647
Last commit
2024-08-06

Live from the GitHub API · updated 2026-10-15

What is Real-ESRGAN?

Real-ESRGAN is the widely used super-resolution model behind many upscalers. It ships portable NCNN/Vulkan executables for Windows, macOS and Linux as well as a Python package, with models for general photos and anime.

Key features

  • General and anime-specific upscaling models
  • Portable executables — no Python needed
  • Optional face enhancement via GFPGAN
  • Video frame upscaling

System requirements

Platforms
Windows · macOS · Linux
GPU
Vulkan GPU for the portable build; NVIDIA for the PyTorch version

How to install Real-ESRGAN

Summarized from the project’s README. Projects change quickly — check the GitHub page for the latest instructions.

Portable executable

  1. 1
    Download realesrgan-ncnn-vulkan for your OS from GitHub Releases and extract it.
  2. 2
    Upscale an image.
    ./realesrgan-ncnn-vulkan -i input.jpg -o output.png -n realesrgan-x4plus

Python

  1. 1
    Clone and install.
    git clone https://github.com/xinntao/Real-ESRGAN.git
cd Real-ESRGAN
pip install basicsr facexlib gfpgan
pip install -r requirements.txt
python setup.py develop
  2. 2
    Run inference on a folder.
    python inference_realesrgan.py -n RealESRGAN_x4plus -i inputs --face_enhance

Strengths

  • Fast and reliable
  • Permissive license

Limitations

  • Can over-smooth fine textures
  • Python install has dependency quirks

Real-ESRGAN alternatives compared

SoftwareStarsLicensePlatformsVRAM
Real-ESRGAN

Practical image and video super-resolution models.

37.1KBSD-3-Clause—
Upscayl

Free desktop app that upscales images with AI.

50.6KAGPL-3.0—
GFPGAN

Blind face restoration for old and low-quality photos.

37.7KApache-2.0—
chaiNNer

Node-based image processing GUI for upscaling and batch pipelines.

6.1KGPL-3.0—
CodeFormer

Robust face restoration with an adjustable fidelity–quality trade-off.

18.2KS-Lab License 1.0 (non-commercial)—
rembg

Remove image backgrounds from the command line, Python or an HTTP server.

25.1KMIT—

FAQ

Is Real-ESRGAN free?

Yes — Real-ESRGAN is open-source software released under the BSD-3-Clause license. Check the license (and any model weight licenses) before commercial use.

Does Real-ESRGAN run on Windows, Mac and Linux?

Real-ESRGAN supports: Windows, macOS, Linux.

What hardware do I need for Real-ESRGAN?

Vulkan GPU for the portable build; NVIDIA for the PyTorch version

How do I install Real-ESRGAN?

1. Download realesrgan-ncnn-vulkan for your OS from GitHub Releases and extract it. 2. Upscale an image.

What are the best Real-ESRGAN alternatives?

Popular alternatives include Upscayl, GFPGAN, chaiNNer, CodeFormer. See the comparison table on this page.

Can I use Real-ESRGAN online without installing it?

If you just want results, AiLoft runs top image models such as GPT Image 2.5 and Nano Banana Pro in the browser — no GPU or setup required.

Skip the install — generate images in your browser

No GPU, drivers or Python needed. Use GPT Image 2.5 and Nano Banana Pro online and compare them side by side.

AI Image Generator
Open-source AI image editing, upscaling & restoration tools