36 formats · zero uploads

The file compressor that
never uploads your files.

nanofile is a free online file compressor for images, PDFs, documents, audio and code, and it runs entirely inside your own tab. Nothing is uploaded, because there is no server here to upload to: drop one file or thirty and we route you to the right tool automatically.

Drop your files here

or browse your device. Batches welcome, we detect every format

0
bytes uploaded, ever
36
formats with real quality
~55%
median reduction across formats
0ms
queue time, it's your CPU
How it works

A file compressor in three steps, none of them a server.

01

Drop your files

We read the extension and the magic bytes, so a mislabelled file still routes correctly. Batches are fine.

02

We route you

You land on the right compressor with everything already loaded. A mix of formats, or a format that is not live yet, goes to Create ZIP instead of a dead end.

03

Compare and save

Drag the preview divider to check the result, tune the settings, then save locally, one file or the whole queue.

Your files never travel.

Every compressor here runs inside your tab. There is no upload endpoint to breach, no retention policy to read, and no queue to wait in. Pull the plug on your wifi and it still works.

  • No upload endpoint exists, so there is nothing to intercept.
  • Works with your network disconnected after page load.
  • Nothing to sign up for, nothing stored, nothing to delete later.
  • No accounts, no watermarks, no daily limits.
  • Analytics count page views and compressions, never file names or contents.
Format coverage

One file compressor, every common format.

Every format below is live, and every one is linked. Nothing in this grid is a promise.

Images
50-75% typical

Instant. Re-encoded natively, nothing to download first.

Documents
40-85% typical

Scans and image-heavy decks shrink hardest. Text stays selectable.

Vector
50-80% typical

SVG is cleaned and rewritten. EPS is stripped of its screen preview.

Code and markup
30-70% typical

Minified, then served smaller again once your host gzips it.

Audio
40-70% typical

Bitrate and channel control. Video stays out of scope on purpose.

Archives
create or shrink

Create ZIP packs files or a folder. ZIP rebuilds an existing archive. This is not an unzipper.

Fonts, ebooks and 3D
50-95% typical on full fonts

Subset the glyphs you render. WOFF2 is the web container; TTF, OTF and WOFF stay in their own format. Ebooks and meshes use their own pipelines.

Convert

Wrong format? Change it here.

For when something will not open a file, or will not accept one. Same rules as everything else: it happens in your tab, and nothing is uploaded. More pairs are coming.

Image conversions
metadata dropped

Turn a picture into a format the thing you are using will actually open. Same rules: it happens in your tab and nothing is uploaded.

PDF conversions
nothing lossy on the way in

A picture into a one page PDF, for the form that will only take one, and back out again. Nothing lossy happens to a JPG or a PNG on the way in.

Font conversions
same outlines, new wrapper

Crossing a font between containers. WOFF and WOFF2 are wrappers, so they carry any font either way. Only OTF to TTF has to refuse a PostScript flavoured font, because that one is an outline conversion rather than a repack.

Audio conversions
tags carried over

Decoded and re-encoded in your tab, so a two hour recording is not stopped by an upload limit. Tags cross unless you ask for them to be dropped.

Got video? Use vidshift.io.

Same in-browser, nothing-uploaded approach, built specifically for video. We'd rather send you to the specialists.

Go to vidshift.io →

Questions

How is this file compressor free?

Compression runs on your own machine, so there is no per file cost for us to pass on. No account, no daily limit and no watermark.

How can compression work without a server?

Your browser already has everything needed to read and rewrite these formats, because it does it every time you open a photo or a PDF. We use what is on your device rather than sending your file to ours.

Is this actually private, or just marketing?

There is no upload route in the code. You can watch the network tab: after the page loads, no request carries your file, because your file never becomes a request.

Can I compress a batch of files?

Yes. Drop as many as you like. Files of the same format queue on that compressor. A mix of formats, or a format that is not live yet, goes to Create ZIP, so nothing is silently skipped.

How do I reduce file size for an email attachment?

Most mail providers stop at 20 to 25MB. Images, PDFs, Office documents and audio all take a target size, so you can type 10MB rather than guess at a quality slider, and those are the files carrying the weight in the first place. Code and fonts have no target box because minifying is not a dial you can turn. A folder of mixed files can go through Create ZIP instead.

What about very large files?

The ceiling is your device memory, not our infrastructure. Compression runs in the background, so even a heavy file leaves the page responsive.

Why is video missing?

Because vidshift.io already does it properly, in the browser, with the same privacy model. Duplicating a tool that good would help nobody, so we send you there instead.

Does it cost anything?

No. Compression runs on your machine, so there is no per-file cost for us to pass on.

How do I reduce the MB size of a file?

Drop it in above and you land on the tool for what it actually is. Images and PDFs get a quality slider and a target-size box, documents and audio get a target-size box of their own, and code, SVGs and fonts get a minifier with nothing to tune because there is no quality to trade away. Type a target size in MB or KB where one is offered rather than guessing at a slider, and the highest quality that still fits is chosen for you.

What is the best way to compress files?

Match the method to what is actually taking up the space. A photo or a scanned PDF loses very little to the quality slider each of those tools defaults to, because it is tuned to be hard to see rather than the smallest possible file. Text, code, JSON and fonts compress losslessly instead, by removing whitespace and redundant structure rather than trading away quality that was never there to lose. A ZIP of files that are already compressed, like JPEGs or MP3s, rarely shrinks further, so the win there is compressing the payloads first, then packing what is left with Create ZIP.

What is the downside of compressing files?

For anything on a quality slider, a real one: push it hard enough and you can see it, most easily as blocking or banding in skies and skin tones. Lossless formats such as PNG, code, fonts and ZIP have no such trade, so there is nothing to lose there beyond the negligible time the browser spends doing the work. Every quality slider on this site defaults to a level chosen to be hard to see, and the result panel always shows a before-and-after preview so you can judge it yourself rather than trust a percentage.