The typos that survive a second read
You have read the email twice. The argument is sound, the formatting is clean, and the word "recieve" is sitting in the second paragraph where you have now looked at it three times without seeing it.
That is the error this checker is built for. It carries a curated list of common English misspellings and compares your text against that list word by word. Paste a draft and the familiar typos come back with a correction and a position you can act on in one click.
The size of that list is the honest part of the story, so here it is: 111 entries. A full English dictionary does not fit comfortably in a browser page, and shipping one that half-works would be worse than shipping a short list that works exactly as advertised. This tool knows the mistakes people actually make. It says nothing about anything else, and it does not pretend that silence is approval.
What the list holds, and what being left off it means
The entries are the misspellings that recur in real drafts: wrong doubled letters, vowels transposed, silent letters dropped.
"accomodate" to "accommodate". "neccessary" to "necessary". "seperate" to "separate". "definately" to "definitely". "recieve" to "receive". "occured" to "occurred". "tommorow" to "tomorrow". "wierd" to "weird". "fourty" to "forty". "untill" to "until". "publically" to "publicly". "mispell" to "misspell", a misspelling that is funny precisely once.
Two entries deserve a note. "alot" corrects to the two-word "a lot", because the single-word version has never been standard. "teh" corrects to "the", since every keyboard produces that transposition eventually.
Because the check is a list lookup and nothing more, a misspelling that is not on the list passes in silence. That is the single most important thing to understand about this tool.
| Your text | Flagged? | Why |
|---|---|---|
| "I definately agree" | Yes, to "definitely" | Exact match on the list |
| "I gaurd the door" | No | "gaurd" is a common error but is not on this list |
| "I dont know" | No | "dont" is not on the list |
| "It beleives me" | No | Only "beleive" and "beleived" are listed, not "beleives" |
| "They're going" written as "their going" | No | Both words are spelled correctly, so this is a confusing word pair, not a misspelling |
| "the colour scheme" | No | British and American variants are not treated as errors |
How the matching works, and what follows from it
The checker scans for words made of letters, ignoring case, and compares each one against the list. Several small behaviors follow from that simple design.
Suggestions inherit the capitalization of the word they replace. "recieve" becomes "receive", "Recieve" becomes "Receive", and "RECIEVE" becomes "RECEIVE". The tool keeps the shape of what you typed instead of flattening everything to lowercase.
A hyphen splits a word, so the second half of "well-recieved" is still caught. Numbers and symbols are not part of a word here, which is why a stray digit or an unbalanced bracket will never appear in the suggestions.
There is no context and no grammar. If you typed "their" where you meant "there", both words are spelled correctly and neither is on the list. That class of error is a different problem, covered in the 50 most commonly misspelled words and in the grammar checker.
Apply, ignore, and why the ignore list sticks around
Each flagged word comes with Apply and Ignore word, and the two do different jobs.
Apply writes the correction into your text at that position and re-runs the check. It does not disturb the other suggestions, because the tool verifies that the text at that position still matches the word it flagged before it writes anything. If you edited that part of the draft by hand between the check and the click, the write is skipped and the list refreshes instead. Nothing changes behind your back.
Ignore word drops that word from all future checks on this browser. It survives closing the tab, so a word you have decided is acceptable stays acceptable tomorrow, and it matches on the lowercase form, which means ignoring "Recieve" also silences "recieve". The "Reset ignored words" button clears the whole list at once.
Where that memory earns its place: any document that quotes misspellings on purpose. A style guide, a teaching handout, or a correction notice is full of words that are wrong deliberately, and without an ignore list those words would be flagged on every single pass forever.
Where this checker earns its keep
Proofreaders run it as a sweep after the human passes, not instead of them. It is quick enough to run on every revision, and the findings are concrete: a word, a correction, a position.
People checking text they did not write. Transcribed interviews, subtitle files, and text recovered from a scan all carry the same handful of errors, and those errors cluster in the words this list happens to know. Scanning someone else's text is where a fixed list helps most, because you cannot predict which words they got wrong.
Writers with habitual errors. Most people misspell the same few words forever. If "seperate" and "definately" keep turning up in your drafts, the tool will keep catching them, and applying the same correction a dozen times over a month is a surprisingly effective way to learn the right spelling.
Anyone sending a short, high-stakes message. A cover letter or a client email. The list is short, but it covers the errors that make a reader wince.
Limitations, stated plainly
The tool cannot flag a word it was never told about, and that limit matters more than the rest. "Gaurd", "dont", and "beleives" all pass unnoticed. So does every real word that is wrong in context, and so does every proper noun and piece of jargon.
It does not check grammar, punctuation, or capitalization, and it does not work in other languages, because the list is English. There is no auto-correct mode: a wrong correction applied silently across a long document costs far more than one deliberate click.
An empty report is not a clean bill of health. It means the draft avoided 111 specific misspellings. The grammar checker picks up a different class of error entirely, and neither tool replaces reading your work aloud.
Frequently Asked Questions
Why was a misspelled word not flagged?
Because it is not on the list. "Gaurd" and "dont" are two common examples that people expect to be caught and are not, and "beleives" is missed because only the base forms "beleive" and "beleived" were included. An unflagged word tells you nothing either way.
Why does it keep flagging a misspelling inside my article about misspellings?
Because the word is on the list, and the list does not read context. Click Ignore word and it stops appearing, including after you close the tab. Use the reset button when you no longer need that exemption.
Does it correct the text automatically?
No. Every suggestion has its own Apply button, and the change lands in your text only when you click it. A silent correction that guesses wrong is much harder to find than a suggestion you chose not to take.
Will it catch "their" when I meant "there"?
No. Both are correctly spelled, and this checker compares spellings, not meanings. Homophones and near-identical pairs are a separate problem, and they are why a spelling pass and a read-through are not the same job.
Why did it propose "RECEIVE" in capitals?
The suggestion copies the capitalization of the word it replaces. Lowercase "recieve" becomes "receive", capitalized "Recieve" becomes "Receive", and an all-capitals version stays in capitals.
Read the sentence before you click Apply
The correction is placed at a fixed position, and that position is only as good as the draft that produced it. So build one habit: apply a suggestion, then read the sentence it landed in before moving to the next one. Three seconds per fix, and it catches the case where the list and your meaning disagree.