My archive had lots of files but I only got one ZIP back. Why?

Above 200 extracted files the page stops emitting one download link per file and hands back only the combined bundle, because a results list of several hundred buttons is not usable. Between two and 200 files you get both, the bundle first and then every file on its own line. A single-file archive gets no bundle at all, just the file itself.

Do the extracted files keep their original timestamps and permissions?

No. They arrive as ordinary browser downloads, so each one carries the date your browser saved it rather than the date recorded inside the archive, and any Unix permission bits stored alongside an entry are not applied. Only the path and the contents survive the trip. If original metadata matters for what you are doing, a desktop 7-Zip is the right tool.

Can I pull one file out of a large archive without decompressing the rest?

Not really. The extract pass runs across the whole archive into a virtual folder in memory, and only then is each produced file turned into a download link, so picking one just means clicking one link after the work is already done. 7z archives are also frequently written as one solid block, which is where a lot of the format's compression advantage comes from, and a solid block genuinely has to be decompressed in order anyway.

How large an archive can this handle?

The guard is on the uncompressed total rather than on the .7z file's own size. The listing pass adds up every entry's recorded size and refuses to go further if the sum exceeds roughly one billion bytes, which stops a small archive that expands into an enormous amount of data from filling your tab's memory. Below that ceiling, your device's own free memory is the real limit.

I know the password for my archive. Where do I type it?

Nowhere, and that is by design rather than an omission. This extractor is built with no decryption path at all, so knowing the password does not help here. The check runs during the listing pass, before anything has been decompressed. Open the archive in a desktop 7-Zip, re-save it without a password, and bring that copy back.

What about a .tar.7z, or a .7z with another archive inside it?

Only the outer .7z layer is undone. A backup.tar sitting inside comes back as a .tar download, still packed, and the same goes for a nested .zip, .rar or second .7z. You would need a tool for that inner format to go one level deeper. The extension check looks at the final extension only, so a file named backup.tar.7z is accepted normally.

Extract 7z Files Online Free, No 7-Zip Install, No Upload

Extract a .7z archive right in your browser. See every file inside, then download files individually or all at once. Nothing is uploaded.

🔒 Private by design: your files are processed locally in your browser and never uploaded to any server.

Real 7-Zip, compiled to WebAssembly and running in the tab

The awkward thing about .7z is not that it is exotic. It is that it is the one common archive format with essentially one implementation. Windows and macOS both open a ZIP by double-clicking it; RAR has a handful of third-party openers; .7z still means installing 7-Zip on macOS, and on any Windows build older than the recent 11 releases that added native support. On a managed work laptop, a school Chromebook or a borrowed machine, that is exactly the thing you cannot do.

So this page carries the program instead of asking you to install it. The extractor is 7-Zip’s own source built to WebAssembly, and the compiled binary ships with the site rather than being fetched from someone else’s CDN. It loads the first time you actually extract something, not when the page opens, so arriving here costs nothing. From then on the archive is decompressed by the same code a desktop 7-Zip would run, inside your browser tab, with nothing transmitted anywhere.

A listing pass runs before a single byte is decompressed

The run is two commands, not one. First the archive is listed in 7-Zip’s technical listing mode, which prints a block of key-and-value lines for every entry: its path, its uncompressed size, its attribute letters, and whether its contents are encrypted. That listing is parsed and inspected before the extraction command is allowed to run at all.

Three things are then checked against it. Encrypted entries stop the run. An archive that holds directory entries but no actual files stops it. An uncompressed total over the size ceiling stops it. Only when all three checks pass does the extraction command execute, on the same still-healthy engine instance, writing every file into a virtual folder that is then walked to collect the results.

That ordering is why a bad archive costs you a listing rather than a gigabyte of memory, and why nothing partial is ever handed back. There is no half-extracted state to download.

Opening a .7z on a machine you cannot install software on

Drop a single .7z file onto the box above, or use Choose a file. One archive per run, so results from unrelated archives never end up mixed in one list.
Press 7z Extractor Online, the button carrying this page’s own name.
When the run finishes, download the file you came for, or take the -extracted.zip bundle to get everything in one click.

The bundle is a plain ZIP rather than another 7z, deliberately. It exists so the “download all” click produces something the machine you are sitting at can already open.

Folder trees arrive flattened into the file names

Downloads are a flat list, not a folder browser, so nested paths are folded into the name rather than lost. An entry stored as Photos/2024/beach.jpg becomes Photos_2024_beach.jpg, and if two folders each contained a notes.txt the second one is handed back as notes-2.txt instead of silently overwriting the first in your downloads folder. You can reconstruct the original structure by eye from the names, but you will be doing it by hand.

The -extracted.zip bundle contains those same flattened names, so it is a convenience wrapper rather than a faithful re-packing of the original tree.

The archives this page will refuse, and what to do about each

Password protection is refused in both of its forms. An archive where only the file contents are encrypted still lists its file names, and each encrypted entry is flagged in the listing. An archive where the header itself is encrypted hides even the names, and 7-Zip responds by trying to prompt for a password on standard input; there is no terminal in a browser tab, so that prompt is wired to end immediately and the attempt is caught rather than left hanging.

A file that is not really a 7z archive, or one whose download was cut off partway, fails in the listing pass with an honest error from 7-Zip itself. In every one of these cases what you see on the page is its single generic failure message rather than a specific reason, so if a .7z refuses to open here it is worth checking its size against the source you got it from before assuming the tool is at fault.

Going the other direction, or dealing with a different format, is a different page each time: Create ZIP Files Online bundles loose files into a ZIP that opens everywhere without extra software, Unzip Files Online is the ZIP-format counterpart to this page, and RAR Extractor Online covers .rar. The rest of the archive and file utilities live under dev tools.

7z Extractor Online