Find Duplicate Files in ZIP

This tool reads a ZIP in your browser and finds files whose content is identical, wherever they sit and whatever they are called, grouping them so you can see redundancy at a glance. It fits the moment before you ship or archive a ZIP: a developer consolidating a project export that accumulated copied assets across folders, or someone auditing a deliverable before sending it to a client.

It also suits backup and archiving housekeeping. Long-lived backup ZIPs tend to collect the same document, image, or library file saved into several directories over time. Scanning surfaces those sets and totals the repeated content, which helps you judge whether a cleanup pass is worth doing, and the tool can do the cleanup for you. For data prep, it is a quick way to flag accidental re-exports before a dataset gets passed downstream.

It is equally useful for office and admin work: tidying a shared resource ZIP, or sanity-checking a coursework or submission archive where the same attachment may have been added twice under different names.

Drop or select a single .zip file. The tool opens it locally and compares every file's size and CRC-32, the checksum of its content that every ZIP stores. Files that match on both are reported as a set of identical copies, largest first, with files scanned, identical sets, extra copies and repeated content summarised at the top and each set expanded to show its full internal paths. Empty files are skipped, since every empty file matches every other, and the summary says how many there were.

Separately, it lists file names that appear in more than one folder with different content, such as two different logo.png files. Those are not duplicates, but they are easy to mix up, so they are shown as a note rather than offered for removal.

Two outputs are available. Download without extra copies builds a new ZIP that keeps the first file of each identical set (in archive order) and everything else exactly as it was stored; the original file is never modified. Export CSV saves a report with each set's size in bytes, file count, repeated bytes and paths, plus the same-name list.

Matching on size plus CRC-32 is very reliable for real-world files: two different files would have to share both by chance, which is extremely unlikely. It is not a defence against someone deliberately crafting a colliding file, but that is not the situation this tool is for.

The kept copy is simply the first in the archive. If the copy you want to keep lives in a different folder, for example the one your web pages link to, the downloaded ZIP may keep the wrong path. In that case use the paths shown here with Remove Files from ZIP and choose exactly which copies go.

Also remember that identical content in different places is sometimes intentional: a template copied into several project folders, or a shared image each chapter references by relative path. Removing the extra copies can break those references, so decide per set rather than acting on the headline number.

Start with the summary: check that files scanned roughly matches what you expect in the archive, then open the largest sets first since those drive most of the repeated content. For each set, read the full paths and ask whether each copy is there on purpose.

If you download the copy without extra copies, open it and confirm that the folders and files you rely on are still where you expect, particularly any that other files link to. The exported CSV is a useful worktable for marking decisions if you prefer to remove copies selectively with Remove Files from ZIP.

For official, financial, or legal archives, verify manually before discarding anything: open the files in a set, confirm the kept copy is the correct, current version, and keep the original ZIP as your fallback.