rdfind

rdfind is a command-line duplicate finder for people who care about content-accurate results and controlled cleanup across folders. It suits Linux, macOS, and Windows users comfortable supplying paths and reviewing a results file. Its strongest case is flexible, recursive matching with conservative defaults; its weakest is the need to understand symlinks, ownership, and command-line actions before automating changes.

Overview

rdfind compares file contents rather than relying on filenames, so it can find duplicates even when files have been renamed. It searches directories recursively and can also accept individual files as command-line arguments, making it useful for checking one backup tree, several directories, or a mixed set of paths. The project dates to 2006 and is authored by Paul Dreik.

The program can also help reduce backup-directory duplication, but it is not a graphical organizer or a recycle-bin workflow. Its default run writes a results file without modifying files, a sensible starting point when the scan covers valuable data. Users must decide what to do next.

Key features

Content-based matching

rdfind narrows candidates by file size, then compares first and last bytes and calculates a checksum. It supports MD5, SHA-1, SHA-256, SHA-512, and xxh128, with SHA-1 as the default. It also accounts for inode and device identity, helping distinguish filesystem references that should not be treated as separate content copies. Automatic selection chooses among duplicate files according to the program's rules, rather than requiring a manual choice for every match.

Controlled actions

A dry run prints what an enabled action would do, which is valuable before deletion or link creation. rdfind can delete duplicates or replace them with symbolic links or hard links, but those options change the risk profile substantially. The results file and dry-run output make review possible; they do not remove the need to verify which copy is retained.

Filtering and scope

Options can ignore empty files and impose minimum or maximum file sizes. These filters are practical for excluding noise or focusing a scan on large backup data. Recursive searches work within one directory or across several supplied locations.

Operational cautions

The manual warns that changing directories while a run is in progress can make rdfind brittle. It also cautions that deletion and linking can expose the process to symlink attacks. rdfind does not handle file ownership or permissions and is not recommended across directories with several owners. Those constraints make it a better fit for an administrator controlling a consistent environment than for shared, multi-owner storage.

Pricing

rdfind is free: the rdfind plan costs nothing and is licensed under GPLv2 or later. There is no free trial because the software is already free, and no paid tier, seat charge, quota, or renewal term is given. This suits anyone wanting a no-cost command-line utility, while the trade-off is that convenience features such as a trash workflow are not part of the offering.

Platforms

rdfind is cross-platform on Linux, macOS, and Windows. The official site says most Linux distributions provide packages and gives apt and dnf installation examples. macOS users can install through MacPorts, while Windows users may use Cygwin. That breadth is useful for mixed-device households and servers, though each platform's installation path assumes some command-line familiarity.

Who it's for

Choose rdfind if you need recursive, content-based duplicate detection across directories and want checksum choices, size filters, and a preview before taking action. It is particularly suitable for backup maintenance and technically confident users who can inspect results and manage filesystem permissions.

Look elsewhere if you want a graphical interface, built-in Trash or Recycle Bin handling, or workflows designed around multiple owners and permission boundaries. rdfind's linking and deletion modes demand careful path control and security awareness.

Pros and cons

Pros

• Conservative default: a scan writes results without changing files, allowing review before cleanup.
• Flexible matching: recursive searches, multiple checksum algorithms, and size filters cover varied backup and archive jobs.
• Broad platform reach: Linux, macOS, and Windows support makes the same utility viable across systems.
• Previewable actions: dry-run output shows the consequences of deletion or linking before execution.

Cons

• Command-line focused: supplying paths and interpreting output will not suit users seeking a visual browser.
• Filesystem risk: symlink attacks and brittleness when directories change make destructive or linking modes unsuitable for casual automation.
• No ownership handling: shared trees with several owners are outside its recommended use.
• No Trash support: deleted duplicates do not have a built-in recycle-bin safety net.

Alternatives

For a broader shortlist, see Duplicate File Finders.

• Czkawka is a free, no-ads option for Android as well as Linux, macOS, and Windows.
• dupeGuru is another free choice across Linux, macOS, and Windows for finding duplicate files.
• Anti-Twin is freeware for private use and supports Linux, macOS, and Windows.
• Duplicate Files Fixer offers Android, macOS, and Windows coverage; its free version scans unlimited files but removes up to 15 duplicates at once, while the paid plan is €35.95 a year.
• fclones is a free, MIT-licensed command-line utility for Linux, macOS, and Windows.
• jdupes is free, MIT-licensed software for exact duplicate matching on Linux, macOS, and Windows.
• Space Wizard has a free tier with up to 100 GB per scan and a paid Pro tier.
• AllDup is free for private users and companies, but Windows-only.

Verdict

rdfind is a strong choice for technically capable users managing duplicate-heavy backups across Linux, macOS, or Windows. Its content comparison, configurable checksums, filters, and non-destructive default provide disciplined control at no cost. Choose another tool if you need a graphical cleanup experience, Trash recovery, or safer handling of shared directories and ownership boundaries.