Your text

Result

Start typing above — the results update as you write.

The lexicon and the two checks

248 names are checked: the long weekday and month names come from the platform's Intl data at run time, and the continents, planets, languages, nationalities and countries are a curated list. The second check reads your own text: if a word appears both capitalised mid-sentence and lower-case mid-sentence, the lower-case occurrences are reported so you can pick one form. A capital at the start of a sentence is never treated as evidence.

Names that are also ordinary words are never flagged alone: March, May, August, Mercury, Earth, Danish, Latin, Polish, Welsh, Danish, Polish, Welsh, Chile, China, Georgia, India, Japan, Jordan, Turkey. "The turkey was dry" is a sentence about dinner, so the tool waits for context it cannot manufacture.

This is not a gazetteer. A town, a company, a product or a person outside the list is simply unknown, and the tool says so by listing it rather than guessing at its spelling. Extending the lexicon is a documented change, not a matter of opinion.

The Proper Noun Checker holds 248 documented names — continents, planets, languages, nationalities and countries, plus the day and month names read from the platform's Intl data — and finds both lower-case forms and names your draft writes two different ways.

What is Proper Noun Checker?

Proper nouns are where a mechanical checker can be genuinely useful and genuinely dangerous, because a name the tool does not know is not a mistake. This proper noun checker draws that line clearly. It checks 248 documented names, it reports the names it does not know instead of correcting them, and it publishes the count so the scope can be judged.

The lexicon has two sources. The long weekday and month names are read from the platform's Intl data at run time, so they are never typed into the engine and cannot drift from the names the environment prints. The continents, planets, languages, nationalities and countries are a curated list that is shipped with the tool and counted on the page. A curated list is the honest description: this is a documented set of names, not a gazetteer, and no one should expect a town or a start-up to be in it.

Nineteen of the names are also ordinary English words, and those are never flagged on their own: turkey, may, march, august, earth, china, chile, jordan, mercury, polish, welsh and their relatives. The turkey was dry is a sentence about dinner, and the tool has no way to know better, so it waits for a capitalised context it can trust rather than firing on the word alone. That single decision removes most of the false positives a name list would otherwise produce. A clean run of the case check therefore says something narrow and checkable: none of the 248 documented names was written lower-case.

The second check reads your own document. If a word appears both capitalised mid-sentence and lower-case mid-sentence, the lower-case occurrences are reported and the capitalised form is shown as the documented one. A capital at the start of a sentence is never treated as evidence, because every first word of a sentence is capitalised, and neither is an ALL-CAPS heading. This is the check that catches the real-world problem: a name introduced one way in chapter one and spelled another way in chapter nine.

Some capitalised words will inevitably fall outside the list, and the tool lists them instead of guessing. Up to twenty of the most frequent unknown capitalised words are shown, with their counts, explicitly marked as never corrected. That list is where your own characters, products and places appear, and it is also the honest measure of how much of the draft the lexicon could not see. Name consistency across a long document is the other half of the job, and it is the half only your own text can supply.

Read the findings as two questions rather than one: is this name documented, and do I write it the same way twice? The first is answerable from the lexicon, the second from the report. Where a name appears in two forms, the majority form is usually the intended one, so start there and bring the outliers into line. Copy report and Download give you the findings, the suggested form and the unknown-name list together, which is what you need to settle a naming decision across a long document.

How to use Proper Noun Checker

  1. Paste your draft; documented names and repeated capitalised words are checked as you type.
  2. Look at the findings first: each names the form found, the documented form and the reason.
  3. Read the consistency findings carefully — they are about your own text, not the lexicon.
  4. Scan the unknown capitalised words list and check the spellings by hand; the tool will not correct them.
  5. Copy the report and apply one form for each name throughout the document.

When to use Proper Noun Checker vs related tools

Use this checker when a document repeats its own names and needs them to agree — a report, a manuscript, a set of product pages. Case rules that are not name lists belong to the Capitalization Checker, which covers sentence starts, the pronoun I and everyday name classes.

When the name is a place spelled two ways for a different reason — colour next to color, centre next to center — the Consistent Spelling Checker answers that question, and the Grammar Checker covers the sentence-level checks over the same text.

Privacy & Security

This tool runs entirely in your browser — no data ever leaves your device. There is no server round-trip, no upload, no logging, and no account required. Your input is processed locally using client-side JavaScript and is never stored, transmitted, or accessible to anyone else. When you close the tab, everything disappears.

Frequently asked questions about Proper Noun Checker