Duplicate File Finder: find identical files by content, not name

Every file is read and hashed inside your browser using the Web Crypto API. The SHA-256 digest of each file is computed locally on your device. Only the resulting digest is used to group files. The file itself never leaves your computer. Open your browser network tab while scanning and confirm there are no outbound requests carrying file data.

Two files are reported as duplicates only when their content is byte-for-byte identical. The scanner computes a full SHA-256 hash of every file and compares the digests. Files that have the same name but different content are never grouped. Files that have different names but identical content are always grouped.

Web browsers deliberately cannot delete files on your disk. No web page has permission to remove a file from your file system. This tool tells you exactly which files are safe to delete, then you remove them in Finder, Explorer or your file manager. That is also the safest design: you stay in control of every deletion.

It depends on how much unique data there is. Files with a unique size are skipped instantly because they cannot be duplicates. On typical folders, most files are unique by size and the scan finishes in seconds. For a photo library with thousands of similarly sized JPEGs, expect a few minutes.

Accurate mode hashes the entire file with SHA-256. This guarantees that grouped files are truly identical. Fast mode hashes only the first 256 KB and last 256 KB of files over 10 MB, plus the file size. In practice collisions are astronomically unlikely, but the guarantee is not absolute. Use accurate mode for critical work.

Yes, for exact duplicates. If you exported the same photo twice, or copied a folder, or ended up with photo.jpg and photo (1).jpg from the same source, the tool finds them. It does not find visually similar photos with different resolutions or formats, because those have different bytes. For visual similarity, use a dedicated photo deduplication tool.

Yes. Use the Choose folder button. Your browser walks the directory tree on your device and passes the file list to the tool. Nothing is uploaded. The folder input is supported in all modern browsers on desktop. On mobile it depends on the browser.

Two files might have the same size and the same prefix but differ in a single byte in the middle. Compare mode tells you the exact offset of the first difference, the hex value, the binary value, and the printable character at that position in each file. Useful for binary files, disk images, firmware dumps, and checking that a copy is a genuine copy.

In accurate mode the browser has to hold the whole file in memory while hashing, so the practical limit is your available RAM. Files over about 500 MB may fail on low-memory devices. Switch to fast mode for very large files. Fast mode only reads two 256 KB windows regardless of file size.

Only if those appear as ordinary folders your browser can access. Google Drive, Dropbox and OneDrive sync folders on your disk work if they are mounted as normal directories. Remote FTP or SFTP servers cannot be scanned because the browser has no filesystem access to them.

One row per duplicate file, with columns for the group number, the SHA-256 hash, the file size in bytes, the human-readable size, and the filename. Sort the CSV by group to see exactly which files belong together. This is useful when you want to script the deletion step with a command-line tool.

The tool works in modern mobile browsers, but file selection and folder scanning are more limited on mobile operating systems. iOS in particular restricts folder access from web pages. For large scans, a desktop browser is a better experience.

When two or more files in the same duplicate group have names that reduce to the same base after stripping copy suffixes like (1), copy, final, backup, the tool flags it. This is a hint that the files were probably created by a copy operation rather than coincidence, so you can be confident about which one is the original.

Only if the audio data is byte-for-byte identical. If one file has a different ID3 tag, the file bytes differ and the tool reports them as different. For metadata-aware music deduplication, use a dedicated audio tool.

SHA-256 is one of the standard cryptographic hash functions. The probability that two different files produce the same SHA-256 digest is roughly 1 in 2 to the 256, which is far beyond any practical concern. For all real-world purposes, identical SHA-256 means identical content.

Some browsers limit how many files a web page can hold at once, especially on mobile. If your folder contains tens of thousands of files, split it into smaller folders and scan one at a time. Also make sure the folder is local, not a network share.