Direct answer
A Turnitin score is the headline number on a Turnitin report, and it is actually two different things: the similarity score, which is the percentage of your text that matched sources in Turnitin's repositories, and the AI writing score, which estimates how much of your submission was likely generated by AI [1]. The similarity score is not a plagiarism verdict — it counts matched text, including correctly cited quotations and references [1]. Turnitin only prints a numeric AI percentage when its confidence is high enough; below that threshold it shows an asterisk (*) instead [3].
What Does the Turnitin Similarity Score Actually Measure, and What Number Is Normal?
The similarity score measures overlap, nothing more. Turnitin compares your submission against its internet archive, its publication index, and previously submitted student papers, then reports the share of your text that matched something in those repositories [1]. A 22% score therefore means roughly a fifth of your wording matched a source — not that a fifth of your paper was stolen. Citations, quoted material, and common phrasing all count toward the match unless your instructor has switched on exclusions for quotes, the bibliography, or matches below a set word count [1][2].
There is no universal "normal" number, because no single threshold exists. Departments and individual instructors set their own expectations, and a figure that raises eyebrows in one module may be perfectly routine in another where long quotations are expected [2]. Turnitin's own guidance is that a score of 0% does not prove originality and a score near 100% is only possible when the text is almost entirely duplicated from elsewhere [2]. Judging your work by the headline number alone will mislead you in both directions.
What matters is the match breakdown behind the number. Turnitin's report lists each source, the percentage it contributed, and highlights the exact strings that overlapped, so you can see whether a high score came from a properly cited block quote or from an unattributed paragraph [1][2]. Instructors read the report the same way — they look at what matched and how it was handled, not at the summary figure in isolation [2]. If you want to lower a score legitimately, the fix is paraphrasing in your own words and tightening citations, not deleting references.
- Read the sources, not the number: a 30% score driven by one cited quotation reads very differently from a 30% score spread across unattributed paragraphs.
- Check your instructor's exclusions: quotes, bibliography, and small matches may or may not be counted [1].
- Treat 0% with suspicion: it can signal a paper with no engagement with the literature rather than flawless originality [2].
What Is the Turnitin AI Writing Score, and Why Does It Sometimes Show an Asterisk Instead of a Percentage?
The AI writing score is a separate indicator that estimates the proportion of your submission likely written by a large language model, reported as a percentage and displayed in its own section of the report [3]. It is generated by a different model from the similarity check, so the two numbers are independent — you can have a low similarity score and a high AI score, or the reverse. The report also highlights the specific segments that pushed the AI percentage up, which lets you see whether the flagged passages are your own analytical prose or generic connective text [3].
The asterisk is the part that confuses most students. Turnitin only displays a numeric AI percentage when its detection confidence clears its own threshold; when the signal is weaker than that, it prints an asterisk (*) rather than a precise figure [3]. An asterisk is therefore not a hidden high score and not a clean bill of health — it means the detector did not have enough confidence to commit to a number. Because of how the threshold works, a genuine low-AI document and a borderline document can both end up showing the same symbol [3].
Two practical consequences follow. First, never treat an asterisk as proof that you are safe, and never treat a visible percentage as proof that you are in trouble — both need the segment-level view to interpret [3]. Second, if your institution reports AI scores through its learning management system, the symbol you see there is the same low-confidence signal described in Turnitin's documentation, not a display bug [3]. When in doubt, look at which sentences were flagged and rewrite the ones that read like generic model output.
How Can a Student See Their Own Turnitin Similarity and AI Scores Before Submitting?
Some institutions let students run a draft through a non-repository assignment, which checks the file without adding it to Turnitin's student paper database, so the report you see before the deadline is not the one that counts against you [4]. Availability depends entirely on instructor and administrator settings, so the first step is asking your tutor whether a draft or practice submission slot exists in your module [4]. If it does, use it early — reviewing the report with time left lets you fix over-quoting, thin paraphrasing, and citation gaps while you can still edit [4].
Where draft submissions are not enabled, students commonly turn to an independent pre-submission checking service to preview the same style of report. turnitin0 is an independent service, not affiliated with Turnitin, LLC, and it returns two PDFs per order — a Turnitin AI detection report and a similarity report — for a .docx, .pdf, or .txt file between 300 and 30,000 words and under 20 MB. The check is non-repository: your file is not added to Turnitin's student paper database and reports are not shared with third-party databases, and you can delete files from your account afterwards. Turnaround is under 15 minutes in 98% of cases, with delivery guaranteed within 30 minutes even during queue spikes, and no subscription is required. The service reports 100,000+ reports delivered to 20,000+ students across the US, UK, Canada, Australia, New Zealand, and Ireland, with a 4.9/5.0 satisfaction rating.
Reading your own report before submission changes how you use it. Instead of reacting to a number after the fact, you can open the match breakdown, confirm that every quoted passage is cited, and check whether any paragraph reads as generic model prose [4]. Students who do this consistently report that the exercise improves the writing itself, not just the score, because it forces a paragraph-by-paragraph audit of where their ideas end and their sources begin [4]. Treat the preview as a drafting tool rather than a verdict, and the score stops being something that happens to you at the deadline.
Understanding the two numbers is the easy part — the harder part is seeing yours before your instructor does. turnitin0 gives students the same style of Turnitin AI and similarity reports their professors open in the LMS, delivered in minutes and without adding your draft to any repository, so you can fix what needs fixing while you still have time.
※ Turnitin0.com - Actual Turnitin AI Report Cover, Score, Flag And Similarity Summary
FAQ
Is a Turnitin score the same as a plagiarism verdict?
No. The similarity score is the percentage of your text that matched sources in Turnitin's repositories, and it counts properly cited quotations and references along with anything unoriginal [1]. Only your instructor can decide whether the matches constitute misconduct, which is why the report shows sources and highlights rather than a pass/fail label [2].
What Turnitin score is considered acceptable?
There is no universal threshold. Instructors and departments set their own expectations, and acceptable ranges vary by discipline and assignment type [2]. Turnitin's guidance notes that 0% does not prove originality and near-100% only occurs with almost fully duplicated text, so the match breakdown matters more than the headline figure [2].
Why does my Turnitin report show an asterisk instead of an AI percentage?
The asterisk appears when Turnitin's AI detection confidence falls below its display threshold, so the system declines to print a precise number [3]. It is a low-confidence signal rather than a hidden high score, and it can appear on both genuinely human-written and borderline documents [3].
Can I check my Turnitin score before I submit?
Sometimes. Some institutions allow non-repository draft submissions that check your file without adding it to the student paper database, but this depends on instructor and administrator settings [4]. Where that option is not available, independent pre-submission checking services can return similar AI and similarity reports before your deadline [4].
Does a high AI writing score mean I will be penalised?
Not automatically. The AI indicator estimates the share of text likely generated by AI and highlights the specific segments behind the percentage, so a flagged passage may be generic connective prose rather than evidence of misconduct [3]. The most useful response is to review the flagged segments, rewrite anything that reads like model output, and discuss the result with your instructor [3].