Direct answer
To check a draft for plagiarism and AI writing online, you need a tool that produces two separate outputs: a similarity report showing which sources your text matches, and an AI writing indicator showing what share of your prose reads as machine-generated. Turnitin's own documentation makes clear these are two different measurements, not one score [1]. Most free online checkers only do the first half, and they do it against a much smaller source pool than your university uses [3]. If you want the numbers your instructor will actually see, you have to check against the same engine — which is why students preview their drafts on a Turnitin report before the real submission.
How Do Online Plagiarism and AI Checkers Actually Work?
An online checker compares your document against a database and reports what it finds. For the similarity side, Turnitin's own guidance is explicit that the resulting percentage is the proportion of a paper's text that matches sources in its database — it is not a "plagiarism score," and a high number does not by itself prove misconduct [2]. That distinction matters because students routinely panic over a 30% figure that turns out to be their reference list and three correctly quoted definitions.
The mechanics are worth understanding before you trust any number. Turnitin generates a match overview that lists each source alongside the percentage of your text it accounts for, and it lets you exclude quotations, the bibliography, and small matches so you can see the substantive overlap rather than the noise [2]. A report that shows 24% but drops to 4% once quotes and references are excluded is a very different document from one that stays at 24%. Reading the breakdown, not the headline number, is the actual skill.
The AI side works differently. Turnitin's detector looks at patterns across long stretches of prose and reports the share of qualifying text it believes was generated by AI, which is why it only produces a result for longer documents and supported languages [1]. It is a signal layered on top of the similarity report, not a verdict. Turnitin itself describes the indicator as not always accurate and says it should be read alongside instructor judgement, with false positives possible for formulaic or heavily edited writing [1].
The practical consequence: a complete online check gives you two reports, and you should read them separately. A clean similarity report tells you nothing about AI flags, and a low AI score tells you nothing about whether you lifted a paragraph from a journal your checker cannot see.
Why Do Free Online Plagiarism Checkers Give Different Results From Turnitin?
Because they are searching a different library. Free checkers typically index the open web, while institutional Turnitin accounts compare your work against subscription journals, book content, and a repository of previously submitted student papers [3]. That last category is the one students underestimate most: the essay on your topic written by a student two cohorts ago is in the institutional database and is not on any free website.
The counting rules differ too, and they change the number more than most people expect. Tools disagree on how to treat quoted material, reference lists, tables, and matches shorter than a sentence, so identical text run through two checkers can produce two different percentages without either being "wrong" [3]. This is why comparing your free-checker score to a friend's Turnitin score is meaningless — you are comparing two different measurements, not two opinions about the same measurement.
There is also a structural gap on the AI side. Turnitin's AI indicator is only visible to instructors who have the feature enabled, and students do not see it on their own institutional submissions [1]. So the score you are most anxious about is precisely the one your university account will not show you. That asymmetry is the single strongest argument for previewing your draft through a service that returns the full report rather than a headline percentage.
None of this makes free tools useless. They are reasonable for catching an unquoted paragraph you forgot to cite. They are not a substitute for a pre-submission check against the engine your institution runs [3].
What Should You Do If Your Draft Is Flagged Before You Submit?
Treat a flag as a revision instruction, not a verdict. Turnitin's guidance on academic integrity and AI writing stresses that institutions expect students to be able to explain their process and to disclose AI assistance where it was permitted [4]. If a passage is flagged, the productive move is to rewrite it in your own voice with specific evidence, examples, and citations — not to run a synonym swap and hope the number drops [4].
Start with the flagged passages themselves. Open the report, read the highlighted spans, and ask of each one: can I defend this sentence out loud? If the answer is no, the sentence needs work regardless of what any detector says. Where the flag is a genuine citation gap, add the citation. Where the prose is generic because you leaned on a model to pad a section, replace the padding with something only you could have written — a specific finding, a number from your own data, a counterargument you actually considered [4].
Then re-check before you submit, and keep your drafts. If a report is ever questioned, a version history with dated drafts and your notes is far more persuasive than an explanation offered after the fact, which is exactly why Turnitin recommends early, documented conversation with your instructor [4]. Students who can show the messy middle of their writing process rarely have a problem.
Finally, be honest about the AI question. If your institution permits AI assistance with disclosure, disclose it. If it does not, a rewritten passage is still an integrity problem if the underlying work was generated for you. The goal of checking online is to catch problems while they are still fixable — not to launder them.
If you want the numbers before your instructor sees them, turnitin0 runs your draft through a real Turnitin check and returns both PDFs in one order — the AI detection report and the similarity report, formatted the way they appear in an institutional LMS. Documents must be English, between 300 and 30,000 words, and under 20 MB; most orders finish in 5–15 minutes, and your file is never added to the student paper repository. Over 100,000 reports have been delivered to more than 20,000 students in the US, UK, Canada, Australia, New Zealand, and Ireland.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is checking my draft online before submitting considered cheating?
No — checking your own work for citation gaps and AI flags is a revision step, and Turnitin's integrity guidance treats explaining your process as normal academic practice [4]. What matters is that you fix what you find rather than submit it unchanged.
Why does Turnitin show an asterisk instead of an AI percentage?
Turnitin displays an asterisk when the AI detection result falls below its confidence threshold, meaning the signal is too weak to report as a number [1]. Treat it as a low-confidence result, not a clean bill of health.
Can a free checker tell me my Turnitin AI score?
No. Free tools generally index the open web and cannot see the institutional repository or subscription content Turnitin compares against, so their percentages are not comparable to a Turnitin report [3].
What does a 20% similarity score actually mean?
It means 20% of your text matched sources in the database — not that 20% was plagiarised [2]. Exclude quotes and references in the match overview to see how much of that overlap is substantive.
How long does an online check take?
Turnitin0 returns both PDF reports in 5–15 minutes in most cases, with delivery guaranteed within 30 minutes during queue spikes. Documents must be English and between 300 and 30,000 words.