Does my text leave the browser?

No. The text you paste is analyzed entirely client-side using a small statistics module bundled into the page. There is no upload, no server call, and no logging — you can confirm it by opening your browser's network tab while typing.

What language does the analyzer support?

The tokenizer treats Latin and Armenian alphabets natively and works for most European languages out of the box. For non-Latin scripts that aren't Armenian, raw word counts still work, but the top-word ranking may include common stop-words because the per-language stop-list isn't loaded.

How accurate is the reading-time estimate?

It assumes 220 words per minute, which is the median for adult silent reading on a screen. Treat it as a rough planning number rather than a precise stopwatch — managers reading on a phone are faster, students reading dense material are slower.

Can I use this output in a public document?

Yes — the metrics on the right panel are derived from your text and are yours to keep, copy, or share. There is no proprietary scoring layered on top.

New · browser only

Field notes lab

Paste any draft text. We compute reading time, density, and a word sketch — everything runs in your tab; nothing is sent to our servers.

Suggested order on this page

Paste rough or long text in the main box — it stays on this device only.
Watch the right panel for reading time, token counts, unique tokens, and weighted words.
When a line feels sharp enough, carry it into the structured analyzer (Decision or Context fields). Open analyzer →

Paste text

Stopword filter adapts for English vs Armenian; other languages still get counts.

Editorial overview

Field notes is a small browser-only writing lab. Paste a draft, a journal entry, the email you are about to send, the one-pager you are preparing for a hard conversation — anything you want to size up before it leaves your head — and the page gives you a calm read-out: how long it would take to read, how many distinct words you used, the eight or nine words that dominate the piece, and the tokens that are likely to drive the rest of the meaning.

Nothing leaves your browser. The text you paste is analyzed client-side using a small statistics module bundled into the page; there is no upload, no API call, no logging. You can confirm it by opening the network tab in your browser dev tools and watching the request list while you type.

What the metrics mean

Reading time is computed at roughly 220 words per minute, the average pace for adult silent reading. If your reader is a busy manager, halve it; if it is your grandmother, double it.
Words and unique tokens together hint at vocabulary diversity. A 600-word draft with 80 unique tokens is likely repetitive; the same length with 280 unique tokens probably reads tighter.
Top words is the strongest signal of what the text is actually about. If “feel”, “maybe”, and “sorry” are in your top eight, the reader will hear hesitation more than content. If “deadline”, “number”, and “decision” are there, you have a signal you can build on.

Useful workflows

Before sending a difficult email or message, paste the draft and check whether your top words match what you actually want to say.
When preparing a brief for the analyzer on the homepage, use field notes to make sure your one-paragraph context has enough specificity — fewer than 20 unique tokens usually means the report will be vague too.
For a long journal entry, scan the top words after a few weeks. If the same emotional vocabulary keeps repeating across entries, that is a signal worth bringing to a therapist or coach.

Frequently asked

Does my text leave the browser?: No. The text you paste is analyzed entirely client-side using a small statistics module bundled into the page. There is no upload, no server call, and no logging — you can confirm it by opening your browser's network tab while typing.
What language does the analyzer support?: The tokenizer treats Latin and Armenian alphabets natively and works for most European languages out of the box. For non-Latin scripts that aren't Armenian, raw word counts still work, but the top-word ranking may include common stop-words because the per-language stop-list isn't loaded.
How accurate is the reading-time estimate?: It assumes 220 words per minute, which is the median for adult silent reading on a screen. Treat it as a rough planning number rather than a precise stopwatch — managers reading on a phone are faster, students reading dense material are slower.
Can I use this output in a public document?: Yes — the metrics on the right panel are derived from your text and are yours to keep, copy, or share. There is no proprietary scoring layered on top.

Curated by the Life Decision Engine Editorial Team · Last reviewed May 5, 2026. Our editorial standards.