Direct answer
A similarity detection tool compares your document against large databases of web pages, academic publications, and previously submitted student papers, then reports the percentage of your text that matches those sources. It is a text-matching instrument, not a plagiarism verdict: a high percentage tells you where overlap exists, and a low percentage tells you no significant matches were found in the indexed corpora. What matters is not the headline number but which passages matched and whether those matches are properly cited.
What Does a Similarity Detection Tool Actually Check in a Document?
A similarity detection tool works by breaking your submission into text segments, generating a fingerprint for each segment, and comparing those fingerprints against its indexed corpora. Turnitin's report surfaces three source categories: internet pages, publications such as journals and books, and the student paper repository, which holds work previously submitted to the institution [2]. The report does not read your argument or judge your intent — it only tells you where strings of words coincide with something already indexed.
The Match Overview ranks every source by how much of your document it covers, so the largest match appears at the top and smaller ones follow in descending order [2]. Clicking any source opens the matching passages with your text highlighted, so you can see exactly which sentences triggered the flag rather than guessing from a single percentage. This is the practical value of the tool: it converts an abstract score into a precise, editable list of locations.
Two settings materially change what you see. If quotes and bibliography are excluded from the report, your reference list and block quotations no longer inflate the total, which is why a paper with a long bibliography can show a very different score depending on how the instructor configured the assignment [1]. Small matches can also be filtered out, so a report may hide dozens of one- or two-word coincidences that carry no significance [3]. Always check which exclusions were applied before you interpret a number.
Finally, the tool distinguishes between matching a source and copying from it. Correctly quoted and cited material still appears as a match, because the tool matches text, not attribution [1]. A well-referenced literature review can therefore show a substantial percentage while containing no misconduct at all, and an unreferenced paragraph lifted from a blog can show a tiny percentage while being genuine plagiarism. The report gives you evidence; the interpretation is yours.
What Similarity Percentage Is Acceptable, and How Is It Calculated?
There is no universal acceptable threshold, and any service that quotes one is oversimplifying. The similarity score is calculated as the proportion of your document's text that matches one or more indexed sources, expressed as a percentage of the total word count [3]. Because the denominator is your whole document, the same borrowed paragraph produces a much higher score in a 1,200-word essay than in a 6,000-word dissertation, which is why comparing raw percentages across different assignments is misleading.
Turnitin itself frames the score as a starting point for interpretation rather than a measure of integrity [3]. A report in the high twenties may consist almost entirely of correctly cited quotations and a reference list, while a report in the low single digits may contain one uncited paragraph that constitutes the actual problem. Institutional policies, departmental guidance, and the instructor's own configuration of the assignment all shift the line, so the only reliable rule is to read the matches rather than the number.
What you should actually audit is the composition of the score. If the largest matches are quotations, standard methodology phrasing, or your bibliography, the overlap is expected and defensible [1]. If the largest matches are body paragraphs that you paraphrased closely from a source without attribution, that is the material you need to rewrite or cite before submission. A useful habit is to open the top three sources in the Match Overview and confirm each one is either cited or genuinely your own words [2].
Be equally careful with unusually low scores. A very low percentage can mean original writing, but it can also mean the tool had no access to the relevant corpus — for example, when a comparison against the student paper repository is disabled, or when the source you drew from is behind a paywall and not indexed [2]. A low score is reassuring evidence, not proof of originality, and it should be read alongside your own citation audit.
Can I Run a Similarity Check on My Own Draft Before Submitting It to My University?
In institutional Turnitin, students generally cannot generate a similarity report on demand. Whether you see a score at all depends on your instructor enabling student access to the report for that assignment, and many courses leave it switched off until after the due date [4]. There is also no supported self-submission route inside the institutional system: your submission is tied to a specific assignment, a specific class, and a specific deadline, so a draft cannot be checked in advance without consuming a submission slot or affecting your submission history.
The practical consequence is that most students discover problems too late. If your score appears only after the deadline, you have no opportunity to rewrite a closely paraphrased paragraph, add a missing citation, or reduce over-reliance on a single source. Timing is the real risk here, not the score itself: a report you can act on is worth far more than a report you can only react to.
A pre-submission check solves that timing problem by giving you the same two reports your instructor sees — a similarity report with per-source matches and an AI writing report — on a draft you control. On turnitin0.com you upload a .docx, .pdf, or .txt file (English, 300–30,000 words, under 20 MB) and receive both PDF reports in a single order, typically in under 15 minutes, with most orders completing in 5–15 minutes and delivery guaranteed within 30 minutes in queue spikes. Checks are non-repository: your file is not added to Turnitin's student paper database, and you can delete it from your account afterwards. Turnitin0 is an independent service and is not affiliated with Turnitin, LLC.
Reading a pre-submission report is the same skill as reading the final one. Open the Match Overview, work down from the largest source, and for each match decide whether it is cited, whether the paraphrase is too close to the original, or whether the overlap is unavoidable terminology [2]. Fix the body-text matches first, then re-check. Two passes usually resolve the issues that would otherwise surface after the deadline, and the exercise leaves you with a document whose citations you can defend in a viva or an academic-integrity meeting.
A similarity report is only useful if you can act on it while the draft is still yours to change. That is exactly what turnitin0 gives you: the same report view your instructor opens, delivered while there is still time to edit, cite, and resubmit with confidence.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Does a similarity detection tool check for AI writing too?
No — similarity matching and AI writing detection are separate analyses, and Turnitin reports them as two distinct scores. A document can show 0% similarity and still be flagged for AI-generated text, which is why a complete pre-submission check includes both reports [1].
Why does my similarity score change between two checks?
The score depends on the indexed corpora at the moment of checking and on which exclusions are enabled. Newly published web pages and newly submitted student papers enter the database over time, and toggling quotes or bibliography exclusion changes the total directly [1].
Is a high similarity score automatically a misconduct case?
No. The score reports text overlap, and overlap from correctly cited quotations, standard methodology wording, or a long reference list is expected. Instructors review the matched passages and their citations, not the headline percentage [3].
Can my university see that I checked my draft first?
Not through a non-repository pre-submission check — the file is not added to the student paper database and the reports are not shared with third-party databases. Your institutional submission remains the only record your instructor sees.
What file types and sizes can I check?
Turnitin0 accepts English .docx, .pdf, and .txt files between 300 and 30,000 words, under 20 MB. Each order returns both a similarity report and an AI writing report as downloadable PDFs.