Check your own Turkish
Write or paste a Turkish text. This page holds it up against exactly the rules the site itself teaches: 54 forms and constructions Turkish does not have, 58 spellings the dictionary writes differently, 0 verb forms and 23 false friends. Each finding comes with the page on which the rule is explained. Everything runs in your browser; the text never leaves your device.
Writing it down helps you remember.
What it is measured against
Every rule in this checker is on one of these pages — and the same table checks this site’s own Turkish when it is built. A finding without a source would be an assertion; here are the sources, even without you checking anything.
What this checker can do — and what it cannot
It does not check your Turkish. It checks 135 particular things in it. A text with no findings is not a text without mistakes: tenses, word order, articles and everything there is no rule for here are invisible to this tool. The reverse holds, though: whatever it reports is explained on a page of this site.
Three levels of certainty, and the difference matters. A mistake is a construction this site explicitly rules out — there is nothing to weigh up. A spelling is not a mistake in the sentence but a word everybody understands and the dictionary writes differently: herkez for herkes, herşey for her şey, cok without its cedilla for çok. A note is a question rather than a verdict: roman is a perfectly good Turkish word — it simply means “novel” and not “Roman”, and whoever wrote it may have meant exactly the novel.
The accusative on a definite object, the choice between -di and -miş and the harmony of most words are left out on purpose. They are common mistakes, but none of them can be recognised from a pattern: kitap okudum and kitabı okudum are both right — “I read a book” and “I read the book” — and only the writer knows which was meant. Where a pattern settles it — evda, iki kitaplar, Istanbul — the checker speaks up; otherwise it says nothing.
The word list below is information, not a mark. It shows which levels the words of your text sit at in this vocabulary. A text made entirely of words from the first level can be very good, and one full of words from the top levels very bad.