Install Cardloader and connect your card scanner

Cardloader is a free desktop app for Windows, macOS and Linux. Installing takes a minute; connecting a Fujitsu/Ricoh fi-8170 for bulk scanning takes one more driver.

Download the latest release All releases

Quick start

Download. Download the installer for your system from the latest GitHub release.
Install. Install Cardloader and open it for the first time, following the guide for macOS, Windows or Linux.
Choose an engine. Open Settings and download the free local model, or paste your Ximilar API token for exact matches, prices and grading.
Connect a scanner. Optional: install the driver and connect a fi-8170 or another TWAIN or SANE scanner, or use a watch folder.
Check the scan settings. Scan in ADF duplex at 600 dpi, with multifeed detection off, in small stacks.
Create a session. Click New session, pick a folder or your scanner, choose how to identify and grade the cards, then click Create & run.
Review and export. Watch the results fill in, fix any match with one click and export the session to CSV.

Step 1

Download card scanner software for Windows, Mac and Linux

Cardloader 0.1.0 is free. Download the installer for your system, then follow its guide below:

System	Download	Requirements
macOS	.dmg	macOS 11 Big Sur or later, Apple silicon or Intel
Windows	.exe or .msi	Windows 10 or 11, 64-bit
Linux	.AppImage, .deb or .rpm	x86-64; the .deb needs Ubuntu 22.04, Debian 12 or newer

Older versions and the release notes are on GitHub releases. The installers are built from the open-source code by GitHub Actions.

macOS Windows Linux

Step 2 · macOS

Install Cardloader on a Mac

Open Cardloader_0.1.0_universal.dmg from your Downloads folder.
In the window that appears, drag Cardloader onto the Applications folder.
Open Cardloader from the Applications folder. The first time, macOS says it could not verify that “Cardloader” is free of malware: click Done.
Open System Settings → Privacy & Security, scroll down to Security and click Open Anyway next to the message about Cardloader.
Confirm with your password or Touch ID, then click Open. From now on Cardloader starts like any other app, and you can eject the disk image.

macOS asks because this early build is not yet signed with an Apple developer certificate. On macOS 14 and earlier, right-clicking the app and choosing Open also works. In Terminal, xattr -dr com.apple.quarantine /Applications/Cardloader.app does the same.

Step 2 · Windows

Install Cardloader on Windows

Run Cardloader_0.1.0_x64-setup.exe from your Downloads folder.
If a blue “Windows protected your PC” window appears, click More info, then Run anyway.
Click through the installer. It installs Cardloader for your user account, without administrator rights.
Start Cardloader from the Start menu.

SmartScreen warns because this early build is not yet code-signed. The installer downloads Microsoft Edge WebView2 if the PC doesn’t have it yet (Windows 11 does). The .msi installs for all users and needs administrator rights. The TWAIN scanner helper is included in both.

Step 2 · Linux

Install Cardloader on Linux

Pick one package. Open a terminal in the folder you downloaded it to:

Ubuntu, Debian, Linux Mint (.deb)

$ sudo apt install ./Cardloader_0.1.0_amd64.deb

apt also installs WebKitGTK, which Cardloader needs. Start Cardloader from your applications menu, or run cardloader.

Fedora, openSUSE (.rpm)

$ sudo dnf install ./Cardloader-0.1.0-1.x86_64.rpm      # Fedora
$ sudo zypper install ./Cardloader-0.1.0-1.x86_64.rpm   # openSUSE

Any distribution (AppImage)

$ chmod +x Cardloader_0.1.0_amd64.AppImage
$ ./Cardloader_0.1.0_amd64.AppImage

If it doesn’t start and mentions FUSE, install libfuse2 (on Ubuntu 24.04, libfuse2t64).

Step 3

Set up the local AI model or the Ximilar API

Free local model (offline)

Open Settings → Identification → Local model, accept the model’s license and click download. It is about 330 MB and can be paused and resumed. New sessions then identify cards on your computer, with no account and no internet.

Ximilar API (optional)

For exact matches across variants, confidence scores, prices and grading, copy your API token from the Ximilar App, paste it into Settings and click Test connection. The token is kept in your system keychain.

Collectibles recognition needs a Ximilar Business plan or a trial. Each card costs API credits: one call per recognition (two in Auto mode) plus one per grading. See Ximilar pricing.

Step 4

Connect a Fujitsu / Ricoh fi-8170 scanner

Cardloader drives the RICOH (formerly Fujitsu) fi-8170 directly, and any TWAIN 2.x scanner with a document feeder on Windows or SANE scanner on macOS and Linux. Connect the scanner over USB and install the driver for your system:

System	What to install	How Cardloader talks to it
Windows	PaperStream IP (TWAIN) for the fi-8170	Bundled TWAIN helper, in its own process so a stuck driver never freezes the app
macOS	`brew install sane-backends`	SANE `scanimage`
Linux	`sudo apt install sane-utils` (or your distribution’s SANE package); optionally RICOH’s `pfufs` driver	SANE `scanimage`

To check the scanner outside Cardloader, run scanimage -L on macOS or Linux, or open it in PaperStream Capture on Windows.
The fi-8170’s network port works only through PaperStream on Windows; the scanner does not support eSCL/AirScan.
ScanSnap and other scanners: ScanSnap models have no TWAIN driver. Let their own software save into a folder and create a Watch folder session; Cardloader analyses each image as it arrives.

Step 5

How to bulk scan trading cards: recommended settings

Cardloader’s scanner defaults are tuned for standard 63 × 88 mm trading cards:

ADF duplex, so the front and back of each card are scanned and paired in one pass.
600 dpi, colour, which is enough for collector numbers and grading.
66 × 92 mm frame, slightly larger than the card, so edges are never cropped.
Multifeed detection off: stiff cards trigger false double-feed errors.
Start with small stacks of about 10 cards, and don’t feed sleeved cards or top loaders.

For developers

Build from source

You need Rust, Node.js 22 or newer, the Tauri prerequisites for your system, and CMake with Clang to compile llama.cpp for the local model.

$ git clone https://github.com/ximilar-com/cardloader.git
$ cd cardloader
$ just setup     # npm ci + cargo fetch
$ just dev       # run the app with hot reload
$ just install   # build and install locally

just is optional; a Makefile has the same targets. To build without the local model, run cargo build --no-default-features in src-tauri/. More in the README.

Reference

Update or uninstall Cardloader

To update, download the new version and install it the same way, over the old one. Your sessions, settings, API token and downloaded models are kept.

System	Uninstall
macOS	Drag Cardloader from Applications to the Bin.
Windows	Settings → Apps → Installed apps, then Uninstall in Cardloader’s ⋯ menu.
Linux	`sudo apt remove cardloader`, `sudo dnf remove cardloader`, or delete the AppImage.

Uninstalling removes the app only. Your sessions stay in the data folder below until you delete it.

Reference

Where Cardloader keeps your data

Sessions, thumbnails, scans and grading overlays stay on your computer. Original images in folder sessions are never modified or moved, and deleting a session removes only the files Cardloader created.

System	Folder
macOS	`~/Library/Application Support/com.ximilar.cardloader/`
Windows	`%APPDATA%\com.ximilar.cardloader\`
Linux	`~/.local/share/com.ximilar.cardloader/`

Start scanning your cards for free

Download Cardloader for Windows, macOS or Linux, point it at a folder or your scanner, and identify your first batch in minutes. No subscription, no account.

Download Cardloader Star on GitHub