Add your file
Choose or drop a .gz or .gzip file. You can paste the contents instead.
GZIP Inspector — Check .gz Headers, CRC-32 and Members
Choose a .gz or .gzip file (a .tar.gz works too) and every gzip member in it is listed: magic bytes, compression method, flags, modification time (MTIME 0 is reported as no timestamp rather than as 1970), XFL, the OS byte labelled as the compressor's own claim, the optional header CRC, extra-field subfields, the stored original filename and comment. The stored filename is shown in both its ISO-8859-1 and UTF-8 readings; names containing .., absolute paths, backslashes, control characters or Windows device names are flagged and a flattened replacement name is offered. Each member is fully decompressed inside the page so its CRC-32 and length can be recomputed and compared with the stored trailer, and the real expansion ratio plus any trailing padding or stray bytes are reported. When the decompressed data carries the ustar marker at offset 257, a heuristic hint names the first tar entry. An optional UTF-8 text preview shows up to 100,000 characters, and the full result downloads as a JSON report.
File types: .gz, .gzipProcessing: This browser
or drop a file here
✓ Stays on this deviceMaximum size: 100 MB
How to use GZIP Inspector — Check .gz Headers, CRC-32 and Members
Let it read the file
The file is read and parsed in this browser — nothing is uploaded.
Check the result
Look over what came back, then copy it, download it, or move on.
Want to know how files are handled? Read the privacy notice →
Common questions
How do I open a .gz file in my browser and view what is inside without extracting it?
Select the .gz file and the page lists each member's header, stored filename and decompressed size, all without extracting anything to disk. Switch the payload preview to UTF-8 text to read up to 100,000 characters of a text file; for a .tar.gz the first file name in the archive is shown. To pull files out, you still need an archive program.
Is my .gz file uploaded to a server when I inspect it here?
No. The .gz file is read and decompressed inside your browser tab and its contents are never sent to a server. Decompressed bytes exist only in memory long enough to compute the CRC-32 and length, are not written to browser storage, and disappear when you close the tab.
How can I open a gz file on a Mac or Windows PC without installing 7-Zip or using the terminal?
Open this page in Safari, Chrome, Edge or Firefox and select the file; no software install is needed to see the structure and checks. Archive Utility on a Mac or Explorer on Windows may unpack some files, but they do not show whether the CRC matched or whether the stored name carries a path.
Why does gzip report 'unexpected end of file', and how can I see where the .gz file is cut off?
That message means the file stops partway through a member, usually because a download or copy did not finish. Here it appears as a truncated header or truncated member; in a multi-member file the complete members before the break are still listed, with the index and offset of the incomplete one.
What does a gzip CRC32 mismatch mean, and can I still use the file?
A CRC-32 mismatch means the decompressed bytes differ from the checksum recorded when the file was made, so the data was damaged. The member table shows the stored and computed values side by side. An ISIZE mismatch is reported separately. Do not rely on that member; fetch the file again from its source.
Why does a downloaded .gz file show 'not a gzip file' (bad magic)?
Every gzip file starts with the bytes 1F 8B; if those are missing, the file is something else carrying a .gz name. Common causes are a browser or server that already decompressed it, a zip archive renamed to .gz, or an HTML error page saved instead of the real download.
What is the gz file size limit, and how much decompressed data can be checked?
A single file can be up to 100 MB, all members together may decompress to at most 256 MiB, and up to 20,000 members are read. Decompression stops at the ceiling and discards what it produced; a stored size larger than deflate could ever produce from the compressed bytes is refused before decompressing.
Is a gz file safe to open if the CRC-32 check passes?
A passing CRC-32 only proves the bytes were not corrupted in transfer; it says nothing about whether the content is harmless or genuine. This tool does not scan for malware. Pay attention to a flagged stored filename, since names with ../ or absolute paths can make some extractors write outside the folder you chose.
What is the difference between a .gz file and a .tar.gz file?
A .gz compresses one stream of bytes; a .tar.gz is several files bundled with tar first and then compressed as one gzip stream. The inspector checks for the ustar marker at offset 257 of the decompressed data and, if found, labels it as a likely tar with its first file name and entry type, as a hint rather than a verdict.
Why does a multi-member gzip file show several members, and is each one checked?
Gzip allows several compressed members to be joined end to end, and decompressors output them as one stream. Each member gets its own row with offset, header, CRC-32, ISIZE and decompressed length, plus totals, so a tiny first member is never mistaken for the whole file. Padding or stray bytes after the last member are flagged.