Use the archive extractor
Load a supported archive and select the entries you need. The tool automatically rebuilds a ZIP as the current selection changes; the original archive is not uploaded or modified.
Supported archive formats
ZIP, TAR, plain GZIP, and TAR.GZ/TGZ are supported. A plain GZIP is treated as one file. A TAR.GZ or TGZ is decompressed and then parsed as a TAR archive.
Support for 7Z is limited to raw headers and file data that use the 7-Zip Copy (store) method. Compressed 7Z headers and methods other than Copy, including LZMA and LZMA2, are reported as unsupported rather than extracted. GZIP processing also requires browser support for the built-in DecompressionStream API. RAR and other formats are not recognized.
Select files from the archive
The entry table shows each file or directory row with its path, type, size, and status flags. Directory rows appear in the table but are not selectable for extraction. Use checkboxes to choose files or click Select all eligible; Clear selection removes the current choices.
Use the search field to filter the table by path. Searching does not select entries. Filter chips help you review selected or flagged rows. Include and exclude patterns control which files are eligible: an include pattern such as *.pdf narrows the eligible set, while an exclude pattern such as __MACOSX/* removes matching entries.
Extract selected files
- Drop one archive into the drop zone, or click browse file. Only one archive can be loaded at a time.
- Review the entry table, apply patterns or filters, and select the files you want.
- Resolve any blocking messages, such as unsafe paths, duplicate output paths, or an oversized selection.
- As soon as the archive is ready and at least one eligible entry is selected, the selected entries are read and placed in an in-memory ZIP automatically.
- Choose Download ZIP and, if needed, Download manifest. Both downloads are optional. Use Change file to replace the current archive.
Safety checks and limitations
- Unsafe paths - absolute paths, drive-letter paths, and
..path segments are blocked from extraction. - Duplicate output paths - the build is blocked when selected entries would write to the same output path.
- Password-protected or malformed entries - these are flagged and excluded when they cannot be read.
- Expansion limits - one entry can be up to 750 MB uncompressed, the selected total can be up to 1.5 GB, and the listing is capped at 50,000 entries.
- ZIP64 - ZIP64 archives are read on a best-effort basis for sizes and offsets beyond 32-bit values.
- Unsupported 7Z compression - compressed headers and non-Copy methods are reported rather than extracted. GZIP cannot be processed when
DecompressionStreamis unavailable.
Privacy and file handling
The archive is read and processed in your browser and is not uploaded or modified. Selected entries are held in memory while the ZIP is built. The ZIP is saved to your device when you choose Download ZIP; the manifest is saved only if you choose Download manifest.
What this tool is not
This tool extracts selected entries; it does not repair damaged archives, merge archives, remove OS-generated junk, or scan file content for malware. For read-only analysis, use Archive Inspector. For removing known ZIP clutter, use ZIP Junk File Cleaner.
Frequently asked questions
Which archive formats can this tool extract?
ZIP, TAR, GZIP, and TAR.GZ/TGZ are supported for listing and extraction. Plain GZIP is treated as one file, while TAR.GZ/TGZ is parsed as a TAR archive after decompression. 7Z support is limited to raw headers and file data using the 7-Zip "Copy" (store) method; compressed headers and methods such as LZMA/LZMA2 are reported as unsupported. GZIP also requires browser support for DecompressionStream.
Does this tool upload my archive anywhere?
No. The archive is read and processed entirely in your browser. Nothing is sent to a server, and the original file is never modified or written back to.
Why do I get a ZIP download instead of a folder of files?
Browsers don't let a webpage save many individual files to a folder without repeated prompts or hitting download limits. Instead, this tool rebuilds your selected entries into a single new ZIP that you download and unzip yourself.
What happens to unsafe or duplicate paths?
Entries with an absolute path, a drive letter, or ".." traversal segments are blocked from extraction outright. If two selected entries would land on the same output path, the tool blocks the build and tells you which paths collide so you can deselect one.