How to use the zip archive comparer
- Choose the first ZIP (A) and the second ZIP (B).
- Optionally tick Ignore capital letters to treat Readme.txt and readme.txt as the same path, and List files that are the same to see all rows.
- Read the summary: how many files are the same, changed, only in one archive, or renamed or moved.
- Open the table for paths, sizes and checksums, and download the comparison as CSV.
Formula
Each ZIP stores for every file its path, size and CRC-32 checksum in the central directory
Same path and same size and CRC-32: the file is the same; same path with a different size or CRC-32: changed
A path only in A and a path only in B with the same size and CRC-32 are matched as renamed or moved
What remains is only in A (removed) or only in B (added); folders are counted but not compared
Nothing is unpacked, so large archives compare quickly. A matching checksum and size is strong evidence of identical content, not proof.
Two versions of a small project
Archive A holds readme.txt (hello), src/main.c, src/util.c, docs/guide.md, old/name.bin (200 bytes) and same.txt. Archive B changes readme.txt to hello world, keeps main.c, guide.md and same.txt, renames src/util.c to src/util2.c, moves old/name.bin to new/moved.bin, and adds added.txt. The comparison reports 3 files the same, 1 changed (readme.txt), 2 renamed or moved, 1 only in B and none only in A.
The checksums shown for readme.txt are 3610A686 in A and 0D4A1185 in B, which differ, as do the sizes, 5 and 11 bytes. Comparing an archive with itself reports every file as the same and no differences.
What the comparison tells you
The ZIP format keeps a CRC-32 checksum of each file inside the archive, so the page can compare two archives without unpacking them. That makes it fast and also safe: no file is run or written to disk. Use it to see what changed between two releases or backups, to check that a download matches the original, or to find out why two archives of the same project have different sizes.
To open an archive and save files, use the unzip extractor. To compare the text inside two files, the CSV diff tool works on tables. To find out what kind of file you have, use the file type identifier.
Limits to know
CRC-32 is a checksum for catching accidents, not a security tool. Two different files can in rare cases share both a CRC-32 and a size, so treat a match as strong evidence rather than proof. Different compression settings do not matter, because the checksum is of the original content. File dates are not compared: two archives made at different times can still hold the same files.
Encrypted entries are listed using the checksum stored in the archive, which may not describe the content you would see after decrypting. Archives with more than 5,000 entries are refused. For making a new ZIP from the files that changed, use the ZIP file creator.
Assumptions and limits
- Each ZIP is at most 25 MB and has at most 5,000 entries.
- Paths are compared as written, including folders; case is ignored only if you tick the option.
- Renamed or moved files are matched one to one on size and CRC-32, in path order.
- Folder entries are counted but not compared.
Frequently asked questions
How are the files compared?
By path, size and CRC-32 checksum, all stored in the ZIP directory, so nothing has to be unpacked.
What does renamed or moved mean?
A file missing from one archive that has the same size and checksum as a file missing from the other one.
Can two different files have the same checksum?
Rarely. A CRC-32 catches accidents but is not a security check, so a match is strong evidence and not proof.
Is anything uploaded?
No. Both archives are read in your browser only.