How Much Space Does File Compression Actually Save? (ZIP vs 7Z vs RAR)

ZIP compression typically reduces plain text and source code by 60-80%, office documents by 30-50%, and already-compressed files like JPEGs or MP4s by less than 5%. 7Z's LZMA2 algorithm usually beats ZIP by an extra 5-15 percentage points on compressible content, while RAR lands somewhere between the two. None of these numbers move the needle on a folder of photos or video, because those files are already compressed before you ever touch an archive tool.

Why does compression ratio vary so much by file type?

Every compression algorithm works the same fundamental way: it looks for redundant, predictable patterns in the data and replaces them with a shorter representation. Plain text is full of that redundancy — repeated words, common letter combinations, whitespace, predictable structure. An algorithm can exploit that heavily.

A JPEG photo or an MP4 video has already had that redundancy squeezed out by its own internal compression (JPEG's DCT-based encoding, H.264 or HEVC for video). By the time a ZIP or 7Z tool gets to it, there's little left to find. This is the single biggest factor in how much space you'll actually save, and it matters far more than which archive format you pick.

Typical compression ratios by file type

The figures below are representative, well-reasoned ranges based on how DEFLATE (ZIP), LZMA2 (7Z), and RAR's proprietary algorithm actually behave on different kinds of content — not a specific lab-measured benchmark run against a fixed test set. Real-world results will vary depending on the exact files, but the relative ordering and rough magnitude hold consistently across everyday content.

Treat these as directional ranges, not guarantees for any specific file. A single text file with unusually repetitive content might compress at 90%; a already-optimized minified JavaScript bundle might only hit 45%. What stays consistent is the ordering — 7Z ahead of RAR ahead of ZIP on compressible content, and all three converging toward zero on already-compressed content.

What does this mean in actual gigabytes?

Percentages are abstract until you attach them to a real folder size. Here's what the ratios above translate to for two common 1GB folders — one of source code, one of phone photos.

That's roughly 700MB saved on the code folder by switching from ZIP to 7Z at typical ratios, versus a difference measured in single-digit megabytes on the photo folder. The format you pick barely matters for the second folder; it matters quite a bit for the first.

Why does 7Z usually beat ZIP by a real margin?

ZIP's DEFLATE algorithm, standardized in 1993, uses a relatively small search window (32KB) when looking for repeated patterns and a simpler entropy coding scheme. It's fast and universally compatible, but it leaves compression on the table by modern standards. 7Z's LZMA2, by contrast, uses a much larger dictionary — often megabytes to gigabytes depending on the compression level — letting it find matching patterns far apart in a large file that DEFLATE's small window would simply miss.

This gap is most visible on large, text-heavy, repetitive content: source code repositories, log files, SQL dumps, structured data exports. On a small file, or on content that's already dense with unique information, the gap narrows because there's less long-range redundancy for LZMA2's bigger window to exploit in the first place.

Where does RAR fit between them?

RAR's algorithm is proprietary and closed-source, so exact internals aren't public the way DEFLATE and LZMA2 are, but its real-world behavior is well documented: it generally beats ZIP's DEFLATE on compressible content, though usually not by as much as 7Z does. RAR5 (the current format generation) narrowed that gap further versus older RAR4, but LZMA2 still tends to edge it out on the most compressible content types, like source code and plain text.

For extraction, this doesn't matter much day to day — Unzipr opens RAR archives for free regardless of which generation created them. It only becomes relevant if you're deciding which format to use for creating new archives, which is where the comparison in this piece is actually aimed.

Can compressing a file actually make it larger?

Yes, and this is a genuinely common point of confusion. Every archive format adds a small amount of fixed overhead per file — headers, a central directory listing, checksums for integrity verification. On content that has zero redundancy left to remove (an already-compressed JPEG, MP4, or ZIP), that overhead has nothing to offset it, so the "compressed" result can end up a few bytes to a few kilobytes larger than the original.

This effect is more noticeable with many small files than with one large file, since the per-file overhead compounds. A folder of a few hundred small already-compressed images can produce an archive that's measurably larger than the sum of the originals, purely from that bookkeeping cost, not from any actual failure of the compression process.

When does the format difference actually matter?

The gap between ZIP and 7Z matters most in a handful of concrete situations: transferring genuinely large files or folders where every gigabyte affects upload or download time, backing up text-heavy data you're storing long-term where storage cost accumulates, or squeezing a file under a hard size limit — an email attachment cap, a form upload limit, a messaging app's file size ceiling. In each case, a real 20-30% difference in compressed size can be the difference between fitting under a limit or not, or between a five-minute upload and a seven-minute one.

It matters far less when you're zipping a handful of small documents to send to a colleague, or bundling a handful of photos for convenience rather than size reduction. In those cases the archive's job is really just bundling and organizing files into one unit, not meaningfully shrinking them — and any mainstream format handles that equally well.

A realistic scenario: uploading a project archive with a size limit

Picture uploading a 2.5GB folder of source code, config files, and documentation to a platform with a 1GB upload cap. Compressed with ZIP at typical text-heavy ratios, that folder might land around 900MB-1.1GB — close to the limit, and possibly over it depending on the exact content mix. Compressed with 7Z at a high setting instead, the same folder more typically lands in the 650-850MB range, comfortably clearing the cap with room to spare. This is exactly the kind of situation where the format choice isn't cosmetic — it's the difference between the upload working on the first try and needing to split the archive into parts.

A realistic scenario: backing up a photo library

Now picture the opposite case: archiving 50GB of JPEG photos from a phone backup for long-term storage. Whether you use ZIP, 7Z, or RAR, and regardless of compression level, expect the resulting archive to land within a percent or two of 50GB either way. Choosing 7Z here and waiting through its slower, more exhaustive compression search buys essentially nothing over ZIP's much faster pass, since there's no meaningful redundancy in JPEG data for either algorithm to find. Fast ZIP compression, or even no compression at all (a plain stored archive), is the more sensible choice for this specific content type — see our guide on fast versus maximum compression settings for more on when the higher-effort setting is and isn't worth the extra time.

So which format should you actually reach for?

For text-heavy content you're compressing for storage or transfer efficiency — source code, documents, logs, structured data exports — 7Z's LZMA2 is the strongest default when squeezing out the smallest file matters. For anything you're sending to someone whose software you don't control, ZIP remains the safest universal choice, since every operating system opens it natively without a dedicated tool. RAR is worth keeping in mind mainly as something you'll receive rather than create, given ZIP and 7Z cover the creation side of this decision well between them. If your content is already-compressed media, skip the whole decision — none of the three formats will meaningfully shrink it, so pick whichever is fastest or most convenient and don't waste time comparing compression ratios that won't materialize. For more on the broader three-way tradeoff beyond just file size, see our full ZIP vs RAR vs 7Z comparison.

A note on how these figures were derived

To be transparent about methodology: the percentages in this piece are representative ranges grounded in the well-documented, publicly understood behavior of DEFLATE, LZMA2, and RAR's compression engine — not the output of a specific benchmark run against a fixed, disclosed set of test files. Actual results on your specific files will vary based on exact content, file sizes, and how repetitive the data genuinely is. Treat the numbers here as a reliable way to reason about which format to reach for, not as a guarantee of the exact percentage you'll see on any single archive you create.

Frequently asked questions

How much does ZIP compression typically save? It depends heavily on content type — roughly 60-70% on plain text, 25-40% on office documents, and close to 0% on already-compressed media like JPEGs or MP4s. There's no single universal percentage, since the amount saved is driven almost entirely by how much redundancy exists in the source data.

Is 7Z always smaller than ZIP? Not always, but usually on compressible content, typically by 5-15 percentage points thanks to LZMA2's larger search window. On already-compressed content, the two formats produce results close enough that the difference is negligible.

Why is my ZIP file almost the same size as the original folder? This is expected if the folder is mostly photos, videos, or other already-compressed media — there's little redundancy left for any compression algorithm to remove. See our dedicated explanation of why compressed files sometimes don't shrink for the full mechanism.

Does a higher compression level always produce a smaller file? Yes, a higher level within the same format and algorithm never produces a larger result than a lower one — at worst it ties, never loses, though the gain shrinks toward zero on incompressible content.

Is RAR worth using for creating new archives? Generally not over 7Z if compression ratio is your priority, since 7Z's LZMA2 tends to edge out RAR on the most compressible content types. RAR is more relevant as a format you'll need to open than one you'd choose to create from scratch.

Can I compress a ZIP file again to make it smaller? No — re-compressing an already-compressed archive gains essentially nothing and can make the result slightly larger due to added format overhead, regardless of which tool or format you use for the second pass.

The bottom line

File type, not archive format, is the dominant factor in how much space compression actually saves — but once content is genuinely compressible, the format choice adds a real, measurable difference on top of that. Unzipr supports all three formats so you're not locked into one regardless of what the data favors: ZIP compression and RAR extraction are free, and 7Z compression with adjustable levels — including the higher settings that matter most for the text-heavy content where the format gap is largest — is available in PRO.