File formats

Online File Tools And Whether Yours Uploads

A privacy policy is a promise. A tool that finishes the job with the network switched off is a demonstration.

7 min read

Online file tools come in two designs that share one name. Online tool means two entirely different things and the difference is whether your file leaves your machine. One kind sends it to a server and works on it there. The other sends you the code and works on the file where it already sits. Both are described with the same words. Telling them apart takes about thirty seconds.

That matters more with mail than almost anything else. A mailbox is not one document. It is every document, every address and every conversation somebody has had, in one file. Deciding where that file goes is worth more than a moment's thought.

Two Online Tool Designs Wearing One Name

Both sit at a web address. Both ask you to choose a file. What happens next is where they part.

The file goes to the server

Upload and process
choose file
  file uploads
  server does the work
  result downloads

your file was on
  somebody else's disk

Now somebody else holds a copy, however briefly. Their retention policy applies to it rather than yours.

The code comes to you

Load and process
page loads
  code arrives
  browser does the work
  result saves locally

your file never
  left the machine

The transfer runs the other way. What crossed the network was the program, not your data.

Neither design is dishonest. What causes trouble is that both are advertised as free online tools, so the words on the page tell you nothing about which one you are about to use.

The padlock settles none of this. It means the connection is encrypted while data moves. It says nothing about whether your data moved at all. It says nothing about what happens after it arrives. Three separate questions, one answer.

How the Browser Became Able to Do This

Local processing was not practical for a long time, which is why uploading became the norm. Two things changed that.

The File API gave pages a way to read a file you choose without sending it anywhere. WebAssembly then let real programs run in a tab at close to native speed, so parsers written in other languages could work inside a browser rather than on a server.

Plenty of tools still upload because they were built before this was possible. The design is not always a decision about your data. Sometimes it is just the year the tool was written, which is worth remembering before assuming anything about intent.

Checking an Online Tool Yourself in Half a Minute

You do not have to take anybody's word for this, including ours. The browser will tell you.

  1. Open the network panel before choosing a file. Right click, Inspect, then Network. Leave it open.
  2. Choose your file and watch the list. An upload appears as a single request carrying roughly the size of your file. A ten megabyte file makes a ten megabyte request. It is hard to miss.
  3. Read the size column rather than the names. You are looking for one large outgoing entry. Small requests are the page fetching its own pieces.
  4. Then try it with the network off. Load the page, disconnect, use the tool. Anything that keeps working is doing the work locally, because there is nowhere else left for it to happen.

Disconnecting settles what a policy cannot. A privacy policy is a promise. A tool that finishes the job with the network switched off is a demonstration. It takes less time to run than reading the policy would.

What Each Online Tool Design Actually Costs

Local processing is not free of trade-offs. Any page claiming otherwise is selling something.

UploadedIn the browser
Who holds your fileYou and themYou
Very large filesHandles themLimited by tab memory
Speed on heavy workServer hardwareYour machine
Works without a connectionNoYes, after loading
Depends on the site staying upEntirelyOnly to load
Anything to delete afterwardsOn their side, if they doNothing was sent

The middle rows are the honest cost. A browser tab has far less memory than an installed program, so a very large file can fail in one and open in the other. That limit is real and it moves as machines improve.

Why Mail Files Raise the Stakes

Most file tools handle one document at a time. Mail files do not work that way.

A PST from a working mailbox holds years of correspondence, every attachment anybody sent, the addresses of everyone involved and a good deal about who spoke to whom. Uploading one is a different act from uploading a photograph, even where the tool is entirely trustworthy.

The fileWhat uploading it hands over
One photographOne image
A PDF contractOne agreement and its parties
A mailbox exportYears of correspondence and everyone in it
A shared mailboxAll of that, for people who never agreed to it

The last row is the one that catches organisations. Uploading a colleague's mailbox to a service you have not assessed makes a decision on behalf of everybody who wrote to them. Under data protection rules that is a processor relationship. It needed thinking about before the file was chosen rather than after.

Read next A Backup Is Not an Archive And the Gap Costs Years Where a mailbox export should live once you have decided how to make it.

When Uploading Your File Is the Right Answer

None of this makes uploading wrong. It makes it a decision rather than a default.

SituationReasonable choice
A file too large for a tabUpload, failing that install something
Work needing a large modelUpload, it cannot happen locally
Content that is already publicEither, the question hardly arises
A mailbox, a contract, medical recordsLocal, failing that a provider under contract
Somebody else's dataLocal, unless you have asked them

Local processing is not an argument for never uploading. It is an argument for knowing which one you picked, since at the moment most people cannot tell and the page rarely says.

What This Site Does And How to Check It

Our viewers and converters read files in the browser. Nothing is uploaded and no copy exists on our side, because there is nowhere for one to be.

What You Should See

  • Requests only for the page and its code
  • Nothing outgoing that matches your file
  • The tool still working when disconnected
  • Results saving straight to your machine

What Would Contradict Us

  • One large request as you choose the file
  • A wait that scales with file size
  • The tool failing the moment you disconnect
  • A download link rather than a saved file
Run this against us and against anybody else. The right column is what an upload looks like. It looks the same whoever built the page.

That is a claim like anybody else's, so do not take it on trust. Open the network panel and load a mailbox. Then disconnect and do it again. The second test is the one that means something. It applies just as well to every other tool you use.

Browser capabilities checked against W3C specifications and vendor documentation in August 2026. The verification steps were run in a current browser.

Questions People Ask

7 questions, answered in full below.

How do I check whether a site uploads my file?

Open the network panel before you choose the file, then watch what happens when you do. An upload shows as a request carrying roughly the size of your file. Nothing appearing means nothing left. It takes about thirty seconds and needs no technical knowledge beyond reading a list.

Is browser processing safer than uploading?

It removes one risk rather than all of them. Your file never reaches a third party, so nothing there can be kept, breached or read. The machine you are sitting at is still yours to look after. A page can still be badly written.

Why do some tools still upload?

Because some work genuinely needs it. Anything drawing on a large model, a licensed font library or serious processing power has to happen somewhere with those things. Uploading is a reasonable design. Not saying so plainly is the problem.

What are the limits of working in a tab?

Memory, mostly. A browser tab has far less to work with than an installed program, so very large files can fail where a desktop tool would not. That boundary moves as machines improve. It has not gone away.

Does a padlock in the address bar mean my file is private?

No. The padlock means the connection is encrypted while data travels. It says nothing about whether data travelled at all. It says nothing about what happens once it arrives. Those are separate questions and the padlock answers neither.

Does anything stay in my browser afterwards?

It can. That is worth knowing. Some pages keep working data in browser storage so the tab survives a refresh. Clearing site data removes it. Nothing about local processing implies local storage, which are two different decisions a page makes.

Can I use these offline?

Often yes, once the page has loaded. Trying it is the strongest test there is. If the tool keeps working with the network switched off, no file is being uploaded, because there is nothing left to upload it to.

Sources

Where the figures and behaviour described above were checked.

  1. File API W3C
  2. WebAssembly Wikipedia
  3. Network features reference Chrome DevTools documentation