Compress · Images · PDF · Video · Audio · Docs · Code · Fonts · Archives
Drop any file and it auto-detects the type and crushes it — right inside your browser. No uploads, no accounts, no limits. Your files stay on your device.
or click to browse · auto-detects every format
Compression happens 100% in your browser. We record only anonymous stats — file type & size — never your files or filenames.
The complete compression guide
Compression is the quiet difference between a page that loads instantly and one people abandon, between an email that sends and one that bounces. Here's a practical, format-by-format guide to shrinking any file well — how much you can really save, what quality costs, and when to leave a file alone.
Smaller files load faster, and speed is money: most visitors leave a page that takes more than a few seconds, and search engines rank slow pages lower. Beyond the web, compression is what gets a deck under an email attachment limit, fits more photos on a phone, trims a cloud-storage bill, and makes backups quicker to move. The savings compound — a homepage with ten right-sized images can weigh a fraction of the original with no visible difference.
Your files are processed on your own device and never uploaded. Nothing is stored, seen, or sent anywhere — the work happens and stays inside your browser.
No account, no watermark, no daily cap, no paywall waiting at file number six. Compress as many files as you want, as often as you want.
Images, PDFs, video, audio, Office documents, code, fonts and archives — all in one place, so you're not hunting for a different site per file type.
Drop a whole folder's worth at once. Results come back grouped by format in ready-made ZIPs, so a hundred files is as easy as one.
Lossless compression makes a file smaller while keeping it byte-for-byte identical in what it produces — used for code, fonts, ZIPs and text. Lossy compression throws away detail your eyes or ears won't miss to save far more — used for photos, video and audio. The trick is picking a quality level where the savings are huge but the loss is invisible. That sweet spot is where a good compressor lives, and it's the default here.
Images are usually the heaviest thing on a web page and the biggest, easiest win. Most photos carry far more data than the screen can show — extra colour precision, hidden metadata, and inefficient encoding. Right-sizing them is the single most effective speed improvement you can make.
JPEG is the workhorse for photographs. Re-encoding at a smart quality level typically cuts 30–70% with no difference the eye can catch, because it discards fine detail selectively. For the web, converting a JPEG to WebP — a newer format built for photos — usually saves another 25–35% at the same visual quality, and every modern browser supports it. If you enable format conversion, that switch happens automatically.
PNG is for sharp graphics, logos and screenshots that need crisp edges or transparency. Reducing an oversized PNG's colour palette to only the shades it actually uses can shrink it 40–80% while looking identical. SVG icons and logos are text under the hood, so they shrink by stripping editor cruft and redundant precision — smaller and still razor-sharp at any size. GIF animations are optimised frame by frame; for anything longer than a short loop, a video format will always be dramatically smaller.
Modern iPhones save photos as HEIC, and cameras and scanners often produce huge TIFF or uncompressed BMP files. These are read and re-encoded into an efficient, universally viewable image — often the most dramatic savings of all, since the originals are so heavy to begin with.
A PDF's size almost always comes down to one thing: the images inside it. A contract that's really typed text is small and hard to shrink further; a scanned document or an image-heavy brochure is enormous because every page is effectively a photo. Knowing which kind you have tells you what to expect.
The first is a clean rebuild: the document is reconstructed to remove duplicated and wasted internal data while keeping text fully selectable and searchable. This is safe and lossless, but the win is modest on documents that are already lean. The second is deep image recompression, which re-encodes the picture on every page at an efficient quality — the only thing that meaningfully shrinks scans and photo-heavy files, where most of the weight lives.
Deep compression treats each page as an image, which is exactly why it shrinks scans so well — but it also means text on those pages becomes part of the picture rather than selectable characters. For a scanned form or a visual brochure that's perfect. For a text contract you need to copy from or search, the clean rebuild is the better choice. The tool automatically keeps whichever result is smaller, so you never end up with a bigger file than you started with.
Video is the heaviest file type by far, and also where compression pays off most. Footage straight from a phone or camera is recorded at a generous data rate for editing, not for sharing. Re-encoding it to an efficient modern format can shrink it dramatically while looking the same on a normal screen — a phone clip or screen recording often drops by well over half.
Where your device has a dedicated video encoder built into its graphics hardware, the tool uses it, which speeds the work up considerably. Even so, video is real number-crunching: as a rule of thumb, a clip takes roughly as long to compress as it does to play. Keep the tab open and in focus while it runs, and start with shorter clips to get a feel for the results.
Uncompressed audio is startlingly large — a few minutes of WAV or AIFF can be tens of megabytes, because it stores every sample in full. Converting to an efficient format at a sensible bitrate keeps the sound essentially indistinguishable to most ears while cutting the size by 80–90%. That's the difference between a voice memo you can email and one you can't.
Bitrate is the dial that trades size for fidelity. For speech, podcasts and voice notes, a lower bitrate is plenty and the savings are largest. For music you want to keep pleasant, a higher bitrate preserves the detail. Files that are already compressed — an existing MP3 or AAC — won't shrink much further without an audible drop, and that's expected: the heavy lifting was already done when they were created.
A modern Office file, an OpenDocument file, or an e-book is really a bundle of parts zipped together — and the parts that make it big are almost always the pictures inside. A wordy report is small; a slide deck full of screenshots and photos is not. Compression works by shrinking those embedded images in place, so the document opens and edits exactly as before, just lighter.
Presentations tend to gain the most, because they're the most image-heavy — a deck packed with photos can shed most of its weight. Word documents with embedded pictures or logos shrink well too. A spreadsheet that's mostly numbers is already compact and won't change much. Older-style files that aren't bundled the modern way, and plain text, have nothing extra to remove — leaving them unchanged is the honest result, not a failure.
Source code is written for humans — spaced out, commented, generously named. Browsers don't need any of that. Minifying removes the whitespace and comments and tightens the code so it does exactly the same thing in far fewer bytes. It's completely lossless: the behaviour is identical, the file is just smaller, so every visitor downloads less and your pages load faster. JavaScript and CSS commonly shrink 30–60%, and JSON and HTML tighten up too.
Web fonts are a hidden weight on many sites. A desktop TTF or OTF font converted to WOFF2 — the format made specifically for the web — typically loses 40–65% of its size with the exact same letterforms. Since fonts load before text can render in them, a lighter font file means text appears sooner. If you build websites, serving WOFF2 is one of the easiest speed wins available.
A ZIP isn't always as small as it could be — how hard it was squeezed depends on whatever created it. Here, a ZIP-family archive is opened, its contents are packed again as tightly as possible, and any images inside are recompressed along the way, so the whole thing comes out smaller while staying a perfectly valid archive. Single .gz files are re-packed at maximum strength too.
If an archive is already full of things that are themselves compressed — photos, videos, MP3s, or another archive — there's little left to gain, because that data was already squeezed. And some container types (like RAR, 7z, TAR and disc images) can't be safely repacked in the browser, so they're read and passed through untouched rather than risk corrupting them. An honest "kept as-is" beats a broken file.
No. Every file is compressed directly inside your browser on your own device. Nothing is uploaded, stored, or transmitted anywhere — your files never leave your computer, which is what makes this both private and instant.
It's genuinely free with no catch: no account to create, no watermark added, no daily limit, and no paywall after a few files. Because compression runs on your device rather than on a server, it costs nothing to run, so there's nothing to charge you for.
It depends on the file. Code, fonts, text and archives are compressed losslessly — identical quality, just smaller. Photos, video and audio use lossy compression tuned to a level where the savings are large but the difference is invisible or inaudible. You can always keep your original and compress a copy.
Yes. It runs in any modern mobile browser, so you can compress images, PDFs and documents on the go. Very large videos are better done on a laptop or desktop, simply because compressing them is demanding work and a computer has more power to finish quickly.
Usually because it was already well optimised — an existing MP3, a photo that's already compact, or an archive full of media has little left to remove. In those cases the original is kept as-is so you never end up with a larger file. That's expected, not an error.
Yes. Add as many files as you like in a single batch, mixing types freely. When they're done you can download each one, or grab ZIP bundles grouped by format — one click for all your images, another for all your PDFs, or everything together.
Your files themselves are never seen or stored. The only thing recorded is anonymous, aggregate statistics — the type and size of files compressed — with no filenames and no content. It's used to understand overall usage, nothing more.
WebP is the best all-round choice for the web: it's built for photos, supported by every modern browser, and typically noticeably smaller than an equivalent JPEG. Enable format conversion and photos are switched to WebP automatically. Keep PNG only where you need transparency or perfectly crisp edges.
There's a generous size limit to keep your browser responsive, and very large files — long, high-resolution videos especially — will take longer since the work happens on your own device. For big videos, keep the tab open and let it finish.
All processed in your browser — nothing is uploaded.