floi

Distraction-free writing, measured

Published 2 September 2026. Nine writing surfaces, counted the same way on the same afternoon.

Free toolNotepadA quiet place to write, with every note kept in this browser. Nothing is uploaded and nothing asks you to sign up first.Open the notepad

The short answer

Distraction-free writing is not a feature you buy. It is the absence of things, and the cheapest way to get it is to notice what is already on your screen and remove it.

Because that is unsatisfying advice, I counted. Nine writing surfaces that all describe themselves as simple, clean or distraction-free, measured on what they actually put in front of you. The spread is wider than any of their landing pages would suggest: from 3 controls and one other company, to 37 controls and seventy.

What I measured, and how

Chrome 151 on macOS 26.6.2, an Apple M1 Max, 2 September 2026. No ad blocker, no extensions beyond the one driving the browser. Each page was opened at the URL that ranks for online notepad, left to settle, then read with a script.

FigureExactly what it counts
RequestsEntries in performance.getEntriesByType('resource'). Cache-independent, because a cached resource still gets an entry.
Other companiesDistinct third-party origins in that same list. Google Fonts counts. So does an analytics beacon.
Controls on the first screenButtons, links, form fields and editable regions bigger than 4 by 4 pixels whose top edge falls in the first 900 pixels, excluding anything hidden by CSS.
Ad slotsins.adsbygoogle, DoubleClick and googlesyndication frames, and aswift_* frames.

There is no byte count in the table, and that is deliberate. Several of these had been opened earlier the same day, so their caches were warm and their transfer sizes were not comparable with one another. A number I cannot compare fairly is worse than no number.

The results

Sorted by requests. The last column is whether you can type on the page that ranks, or whether it is a landing page with a button on it.

SurfaceRequestsOther companiesControlsAd slotsType here?
floi Notepad150120Yes
Protected Text16130No
RapidTables Notepad327260Yes
Notepad (js.org)362160Yes
Online Notepad .net399104No
Memo Notepad47490Yes
Online Notepad .org53103316Yes
Online Notepad .io1487370Yes
aNotepad17370202Yes

floi's own notepad is the first row, and the reason it wins on two columns is that it is newer than the others rather than cleverer. It also loses on features to at least one of them, which is further down.

Seventy other companies

The number in that table I did not expect is aNotepad's. Opening the page contacts 70 distinct third-party origins across 173 requests, before you type anything. They are advertising identity and bidding services: ID5, Criteo, PubMatic, OpenX, Rubicon, Index Exchange, Amazon, GumGum, Quantcast and a long tail behind them.

None of that reads your note. It is building an identifier for you, which is a different thing and a more durable one. It is worth holding next to the other fact about that site: your note is saved to their server and is Public by default unless you register an account.

The general rule is duller than the number. A page you type private things into should be contacting nobody. Any figure above zero in that column is a page where somebody other than the site owner knows you were there.

The thing I expected to find, and did not

I went in assuming the story would be lag. A heavy editor should stutter on a long document, and the cursor falling behind your fingers is the most physical distraction there is.

It is not the story. I loaded a 5,000-word document into each editor built on a plain text box and timed the site's own handling of one more keystroke, twenty-one times. Every one of them came in at a median of 0.0 to 0.2 milliseconds. A browser text box is very good at being a text box, and none of these tools does enough work per keystroke to matter.

That result is worth publishing precisely because it kills a plausible theory. The difference between these tools is not speed. It is entirely what they put on the screen and what they do with your text afterwards.

What happens when you reload

Separately from the counting, I typed a sentence into each of the seven browser notepads and reloaded the page.

  • RapidTables came back empty. Its own banner warns you about this, which is more honest than most, but the note is gone. Its localStorage held 45 keys after one visit. All of them were advertising identifiers and none of them was my text.
  • onlinenotepad.org came back holding the wrong draft. It saves through TinyMCE's autosave plugin, which writes on a timer rather than when you pause. On reload it restored a draft written 27 minutes earlier, and the sentence I had typed five seconds before was never saved at all. I could not pin the exact interval from a background tab, so I am not printing a number for it.
  • The rest kept it. onlinenotepad.io in IndexedDB, notepad.js.org and floi in localStorage, aNotepad on its own server.

The lesson generalises past these nine. Browser storage is a convenience, not a backup, and it goes when you clear site data. Safari also clears it after seven days of not visiting the site. Anything you would be upset to lose belongs in a file.

The options this page is not selling you

A browser tab is one answer and it is not the best one for everybody. These are the others, with what they actually cost.

OptionRoughlyWhat it is genuinely good at
Freewrite (Traveler, Alpha)$349 and upBeing incapable of anything else. An e-ink screen, a keyboard and no browser. The decision is made once, at the till.
FocusWriterFree, open sourceA full-screen desktop app with daily goals and timers. Windows, macOS and Linux. The obvious first thing to try before spending anything.
iA WriterAround $50 onceTypewriter mode and focus mode done properly, plus a syntax highlighter for your own weak sentences.
Your operating systemFree, already installedmacOS Focus and Windows Focus Assist silence notifications, which is most of the problem and costs nothing.
A browser notepadFreeBeing one tab away from what you were already doing. Nothing to install, nothing to launch, and the notes stop at the machine.
PaperAlmost nothingStill the only one of these with no software update. Bad at search, backup and legibility.

What to do this afternoon

In rough order of how much they buy you per minute spent.

  1. Turn on the focus mode you already own

    macOS Focus, Windows Focus Assist, or Do Not Disturb on the phone next to you. This is free, takes a minute, and removes the interruptions you do not choose.

  2. Write somewhere with no font menu

    Not because Markdown is superior, but because a formatting toolbar is an invitation to edit while you draft. Those are two different jobs and doing them at once does neither.

  3. Put the editor full screen

    Then use the app's own focus mode on top of it, if it has one. The tab bar is a list of other things you could be doing.

  4. Set a short timer and stop when it stops

    Twenty-five minutes is the usual figure and there is nothing magic about it. The pomodoro timer here runs one, and the guide beside it explains why the number matters less than the stopping.

  5. Only then consider spending money

    If the four free things above have not helped, a $349 device is unlikely to. If they have helped and you want them enforced rather than chosen, that is when the hardware makes sense.

Where floi's own notepad loses

The table above puts it first on two columns, so here is the other side of it. notepad.js.org is a better tool than floi's on features and it is not close: it exports PDF, HTML and DOCX, it installs as an offline app, it has a writing timer, snapshots, break reminders and twenty ambient sounds, and it is open source. floi's writes plain text and nothing else.

Its one real cost is the 235 words of instructions it types into your editor for you to delete before you start, and two analytics origins. If those do not bother you, use it.

The honest summary

Nothing in the table above will make you a writer, and most of what is sold as distraction-free is a landing page with an ad slot on it. What the numbers do show is that the range is real: some of these pages are a text box, and some contact seventy companies before you have typed a word.

Pick one at the quiet end and stop reading about it. That last part is the step everybody skips.

Frequently asked questions

Does distraction-free writing actually make you write more?

Nobody has shown that, and the studies usually cited do not say it. What the attention research does support is narrower and duller: switching costs are real, and every switch has a resumption lag before you are back at full pace. A plain editor does not make you productive. It removes the objects on screen whose entire job is to invite a switch, which is a smaller claim and a true one.

Is a dedicated writing device worth the money?

Only if your problem is genuinely the internet rather than the writing. A Freewrite is around $349 for a machine that does one thing, and the people who love theirs mostly describe the same mechanism: it cannot open anything else, so the decision is made once at purchase rather than forty times a day. If you would answer email on it given the chance, it will not help. That is a question about you, not about the hardware.

What is the difference between focus mode and full screen?

Full screen hides the browser or the operating system. Focus mode hides the application's own furniture: the sidebars, the toolbars, the word count, sometimes every line except the one you are on. Most editors that offer both let you stack them. The second one is the one that matters, because the toolbar you keep glancing at belongs to the app you are writing in.

Is typewriter scrolling worth turning on?

It is worth trying for a week. Typewriter scrolling pins the line you are typing to the middle of the screen and moves the text underneath it, so your eyes stop drifting down the page. People either stop noticing it within a day or find it faintly seasick, and there is no middle group. It is in iA Writer, Ulysses and several others, usually under Typewriter Mode.

Should I write in Markdown or in a word processor?

Write in whichever one you will not fiddle with. Markdown wins on this by accident rather than design: there is no font menu, so there is no font to change at paragraph three. The plain text also outlives the tool, which matters more than it sounds like it does. If the document has to end up in a specific layout, the honest answer is that you will paste it into a word processor eventually either way.

Do the browser notepads keep what I write?

Mostly, and not reliably. I typed into all seven that rank and reloaded them. One came back empty. One came back holding a draft from 27 minutes earlier and had never saved the sentence typed five seconds before. The rest kept it. There is a table above with the details, and the general rule holds regardless: if losing it would hurt, it should be a file on your disk rather than a row in a browser store.

Word and character counterFor when the length is the question, with reading time and page count.Pomodoro timerStep four, running in a tab. It survives a reload, which four of six rivals do not.How long should a pomodoro be?Why the twenty-five is arbitrary and the stopping is not.PrivacyThe zero in that third column, and how it is kept at zero.

↑↓ move↵ openesc close