Upscayl Guide: Features, Hardware Needs, and How to Get Started

What it is

Upscayl is an open-source software application designed to enlarge and enhance low-resolution digital images. Created by developers Nayam Amarshe and TGS963, the project describes itself in its repository documentation as a free and open-source artificial intelligence image upscaler compatible with Linux, macOS, and Windows.

In traditional image editing software, enlarging a picture typically involves standard mathematical interpolation. Interpolation algorithms work by calculating new pixels based on the colors of surrounding pixels and stretching the existing image across a bigger canvas. This conventional method often results in blurry lines, smoothed textures, or pixelated artifacts when graphics are scaled up significantly. In contrast, according to the project README, Upscayl uses deep learning algorithms to predict and reconstruct missing details, enlarging pictures without losing quality.

Under the hood, Upscayl relies on neural network research from the Real-ESRGAN project (developed by researcher Xintao Wang in 2021) and uses the Vulkan graphics architecture to accelerate processing on local computer hardware. While many users interact with the graphical desktop application, the project ecosystem also includes a dedicated command-line interface (CLI) known as upscayl-ncnn. This backend engine performs the underlying image processing tasks and is open-source under the GNU Affero General Public License version 3 (AGPLv3). In addition to its desktop tool, the project also provides Upscayl Cloud, a web-based service that offers cloud-powered upscaling and image tools across desktop, tablet, and mobile devices.

What it can do

According to the project documentation and official website, Upscayl includes a comprehensive set of features tailored to restoring images, expanding dimensions, and running batch operations:

• AI-Powered Upscaling: The official website states that Upscayl can enlarge image resolutions by up to 16x. The neural network algorithms assess pixel patterns and guess what fine details should look like, sharpening textures and edges rather than merely stretching existing pixels.
• Multiple AI Models: Different kinds of visuals require different processing approaches. Upscayl provides multiple pre-configured models for various content types. Documented models include High Fidelity (based on the HFA2k model by Helaman), Remacri (created by Foolhardy), and UltraSharp together with Ultramix Balanced (developed by Kim2091). These models allow users to pick specialized weights suited for photography, digital illustrations, or artwork. The project also documents support for converting and importing custom community models.
• Batch Processing: Through its “Batch Upscayl” feature, users can queue and process multiple images in a single run. The project documentation notes that when a chosen model does not support a specific action immediately, the program processes all upscale passes first before finishing secondary post-processing tasks.
• Double Upscayl: For situations where an image needs extra resolution, the software provides a double-pass feature. This setting runs the AI enhancement algorithm over the image twice consecutively to expand dimensions even further.
• Color Fidelity: The official website notes that Upscayl is built to preserve accurate colors and intricate details, making it useful when preparing files for physical printing.
• Cloud Tools: Beyond the standalone desktop application, the official website highlights cloud capabilities on Upscayl Cloud. These include unlimited cloud storage, image editing with facial enhancement and detail recovery, and an AI image generation tool that creates visual styles from text descriptions.
• Processing Speed: According to the official website, the platform’s models are optimized to deliver results up to ten times faster than earlier versions when running on supported acceleration architectures.

The project README also outlines clear boundaries regarding image quality. Upscayl is specifically engineered to improve low-resolution and pixelated imagery; however, it cannot perform optical de-blurring or fix pictures that are completely out of focus.

How to get started

Upscayl can be installed across Linux, macOS, and Windows operating systems, or compiled directly from source code by developers.

On Linux distributions, Upscayl is packaged in several standard formats, including Flatpak via Flathub, Snap packages via Snapcraft, the Arch User Repository (upscayl-bin), and Gentoo ebuilds. Installation files are also available as DEB packages for Debian and Ubuntu, RPM packages for Fedora, and standalone ZIP archives for x86 architectures. For a portable installation, users can download the AppImage:

1. Download the upscayl-x.x.x-linux.AppImage package from the official website or repository releases page.
2. Right-click the downloaded AppImage file, open the permissions properties, and enable the option to allow the file to execute as a program.
3. Double-click the AppImage file to run the application.

On macOS, the software requires macOS 12 or newer. Users can download the DMG archive (upscayl-x.x.x-mac.dmg), open it, and drag the Upscayl icon into the Applications directory. Because the package is downloaded outside the standard store channel, the documentation explains that users should open Finder, right-click Upscayl inside Applications, select “Open,” and confirm the security prompt. Upscayl is also available via the Mac App Store and through Homebrew with the following terminal command:

brew install --cask upscayl

On Windows, the application requires Windows 10 or later. Users can download the installer file (upscayl-x.x.x-win.exe) and double-click to run setup. If Windows SmartScreen displays an unverified publisher warning, click “More Info” and select “Run Anyway” to complete the installation.

For developers who want to run or package the desktop application locally, the project documentation suggests installing Node.js via Volta using:

volta install node

Once Node.js is ready, the source code can be cloned, dependencies installed, and the local development server started with these exact commands:

git clone https://github.com/upscayl/upscayl
cd upscayl

# INSTALL DEPENDENCIES
npm install

# RUN THE DEVELOPMENT SERVER LOCALLY
## YOUR LOGS WILL NOW APPEAR IN THE TERMINAL
npm run start

To create application distribution packages locally, the repository provides the following build scripts:

# INSTALL DEPENDENCIES
npm install

# PACKAGE THE APP
npm run dist

# PUBLISH THE APP, MAKE SURE TO ADD GH_TOKEN= IN SHELL
# ONLY DO THIS IF YOU'RE A MAINTAINER
npm run publish-app

Pricing and requirements

Hardware requirements are an essential consideration when running Upscayl locally. According to the project README, the software requires a graphics processing unit (GPU) that supports the Vulkan API. Vulkan is a modern, cross-platform graphics programming interface that provides direct control over GPU hardware, which the underlying NCNN framework requires for neural network computations.

Because of this Vulkan requirement, the documentation states that Upscayl is not compatible with most central processing units (CPUs) or standard integrated graphics chips (iGPUs). While community-contributed workarounds for non-discrete graphics on Windows and Linux have been documented in the repository’s issue tracker, a dedicated Vulkan-compatible graphics card is the standard system requirement. In setups featuring multiple graphics cards, users can specify which card to utilize via the GPU ID setting. The README notes that Windows users may need to configure system graphics settings to performance mode so the operating system does not override this card selection. Supported desktop operating systems include Windows 10 or later, macOS 12 or later, and modern Linux distributions.

Regarding licensing and commercial tiers, the desktop application and its upscayl-ncnn backend are published as open-source software under the AGPL-3.0 license. For cloud features, hosted processing, and service options, pricing and plans are subject to the official page at upscayl.org.

Who it is for

According to the official project website, Upscayl is built for creators, businesses, designers, artists, developers, and general consumers who work with visual media.

Graphic designers and businesses can use the tool to upscale legacy branding assets, small product photos, and digital marketing graphics while preserving sharpness and color accuracy. Digital artists and illustrators can scale up smaller canvas creations to prepare artwork for high-resolution displays or print portfolios.

The software is also intended for everyday users seeking to restore old family photographs, downscaled digital scans from earlier decades, or compressed web pictures. Because Upscayl Desktop processes all images directly on local hardware without sending files across the internet, it is well suited for privacy-focused individuals and organizations handling sensitive graphical content. Meanwhile, users without a dedicated Vulkan GPU or those working on mobile devices can access the platform via its cloud offering.

Alternatives

ImageMagick is an open-source software suite designed for displaying, converting, and editing raster image files across various formats.

libvips is an image processing library designed to process images efficiently with low memory requirements.

Sources

• https://github.com/upscayl/upscayl
• https://upscayl.org/
• https://github.com/ImageMagick/ImageMagick
• https://github.com/libvips/libvips