Turnitin AI and Similarity Report Services
Send us your assignment, dissertation, thesis or research paper and get the full Turnitin similarity report and the Turnitin AI writing report as PDFs, with a plain English explanation of every match and a practical plan to bring the score down.
- Similarity report PDF
- AI writing report PDF
- Digital receipt
- Source by source breakdown
- Reports in 30 to 45 minutes
The combined total of all matches, including overlapping sources, for each database.
- Bibliography
- Quoted Text
- Cited Text
Illustrative layout only. Not a real student report, and not a promise of any particular score.
What is a Turnitin report?
A Turnitin report is the document Turnitin produces after a file is submitted for checking. It arrives as a PDF and it carries two separate assessments that people often confuse with each other.
The similarity report measures how much of your text matches material Turnitin can see: public web pages, licensed journals and books, and papers submitted previously by students anywhere in its network. It gives an overall percentage, then breaks that figure down source by source and highlights the matching passages inside your document.
The AI writing report is a different check entirely. It estimates how much of the long form prose in your document was likely produced by a large language model, or produced by one and then revised. It has its own percentage, its own limitations, and its own disclaimer printed by Turnitin on the report itself.
Neither number is a verdict. A similarity score counts matched text, including text you quoted and cited perfectly well. An AI score is a probabilistic estimate that Turnitin explicitly says should not be the sole basis for action against a student. What matters is what sits behind the number, which is exactly what we explain when we send your report.
A Turnitin report has two parts. The similarity report shows what percentage of your text matches existing sources and where each match came from. The AI writing report gives a separate estimate of how much of the prose was likely written by AI. The two figures are independent, and on their own neither one proves misconduct.
The two reports you receive
One submission produces two documents. Here is what each one actually looks like and what it is measuring.
The combined total of all matches, including overlapping sources, for each database.
- The overall similarity percentage
- Filters applied, such as bibliography, quoted text and cited text
- Four match groups, sorted by how the matched text was handled
- Top sources split into internet, publications and submitted works
- Integrity flags for anything unusual in the file
- Every page of your document with matches highlighted and numbered
The similarity report is the one most institutions look at first. It is descriptive rather than accusatory: it tells your marker where your words overlap with someone else’s, and leaves the judgement to a human.
The percentage indicates the combined amount of likely AI generated text as well as likely AI generated text that was also likely revised by AI.
- The overall AI percentage for qualifying prose
- A sentence count for text marked as likely AI generated
- Turnitin’s own disclaimer on accuracy and human judgement
- The submission rendered with flagged sentences marked
The AI report is newer, narrower and far more easily misread. It only assesses long form prose, it can misidentify human writing, and Turnitin itself says it takes further scrutiny and human judgement before anyone concludes that misconduct occurred.
What is a good Turnitin similarity score?
Most universities are comfortable below 15 to 20 percent, and many treat anything under 10 percent as unremarkable. There is no universal pass mark. Your handbook, your department and sometimes your individual marker set the threshold, so check yours before you judge your own report.
| Score band | What it usually means | What we normally do about it |
|---|---|---|
| 0 to 10% | Comfortable on almost any course. Matches are typically reference lists, standard method phrases and correctly quoted material. | Confirm the matches are genuinely trivial and send the report with a short note. Usually no editing needed. |
| 10 to 20% | Common and often accepted, particularly in methods and literature chapters where terminology repeats. | Check that every matched passage is either quoted and cited, or paraphrased properly. Tidy the few that are not. |
| 20 to 40% | Needs attention. Usually caused by heavy quoting, patch writing from one or two sources, or a missing citation block. | Work through the flagged passages with a human editor, rewrite what is too close, and fix the citations. Then re-check. |
| Above 40% | Serious. Large sections are tracking closely to one or more sources, or a previous submission of the same work is matching against itself. | Identify whether it is self match or genuine overlap first, then restructure and rewrite the affected sections before resubmission. |
One important exception: if your own earlier draft was submitted to Turnitin before, your new file can match it at a very high percentage. That is a self match, not plagiarism, and it is usually resolved by asking your institution to exclude the earlier submission.
Match groups explained
Turnitin sorts every match into one of four groups. The group matters far more than the headline percentage, because it tells you whether the overlap is a citation problem or nothing at all.
| Match group | What Turnitin means by it | How it is normally fixed |
|---|---|---|
| Not Cited or Quoted | Matched text with neither an in-text citation nor quotation marks. This is the group that causes most academic misconduct queries. | Either rewrite the passage genuinely in your own words and cite the source, or quote it properly with the citation attached. |
| Missing Quotations | The passage is cited, but the wording is still very close to the original and it is not inside quotation marks. | Put the borrowed wording in quotation marks, or paraphrase it far enough that it is your own sentence structure. |
| Missing Citation | The text has quotation marks but no in-text citation, so the reader cannot tell who is being quoted. | Add the author, year and page number in the style your course requires. |
| Cited and Quoted | Correctly attributed material. Turnitin still counts it in the percentage, but it is handled properly. | Usually nothing. The only reason to reduce it is if the document is over quoted and needs more of your own analysis. |
Where the matches come from
Top sources are grouped by database, and each behaves differently.
- Internet sources. Public web pages Turnitin has crawled, including blogs, university sites and study mills.
- Publications. Licensed journals, books and conference proceedings. Matches here are common in literature reviews.
- Submitted works. Papers previously submitted to Turnitin by any institution in the network. A large match here is often your own earlier draft or a file a classmate shared.
Integrity flags
Separately from the percentage, Turnitin flags files that look manipulated. The flags we see most often are hidden text where characters have been coloured to blend into the white background, replaced characters where letters are swapped for lookalikes from another alphabet, and unusual spacing used to break up matches.
These tricks do not work. They are detected, they are flagged in bold on the front of the report, and they turn a similarity question into a deliberate deception case. We never use them and we will tell you if we find them in a file you send us.
Why is my own writing flagged as AI?
AI detection is probabilistic, not forensic. It scores patterns, not authorship, so genuinely human writing can be flagged when it is unusually even, heavily grammar corrected, translated, or written to a rigid template. Turnitin prints this caveat on the report itself.
Common causes of a false positive
- Heavy use of a grammar or rewriting tool, which smooths sentence variation into a machine like rhythm
- Writing in a second language, or drafting in one language and translating into English
- Formulaic academic phrasing that many students are taught to use
- Short, uniform sentences with very little variation in length
- Following a rigid template supplied by the department
What the AI check does not assess
- Bullet lists, tables, code blocks, equations and reference lists
- Documents with too little continuous prose to score reliably
- Images and anything scanned rather than typed
- Intent, which no detector can measure
Because of this, the AI percentage on a heavily tabulated document can look very different from the same content written as continuous prose.
What you receive
Not just a number. Everything you need to understand the report and act on it.
Similarity report PDF
The complete report, including the overview page and every page of your document with matches highlighted and numbered.
AI writing report PDF
The separate AI assessment with the overall percentage, the flagged sentence count and Turnitin’s disclaimer.
Digital receipt
The submission record showing the submission ID, date, file name, page count and word count.
Source by source breakdown
Each match listed with the percentage it contributes and the database it came from.
Plain English summary
A short written note telling you which matches matter, which are harmless and what your marker will most likely notice.
Reduction plan
If the score needs to come down, a specific list of passages to fix and how to fix each one.
Flag check
Confirmation of whether any integrity flags were raised on the file, and what triggered them.
Re-check after edits
Once the edits are in, we run the document again and send the updated report so you can see the change.
Documents we check
Any academic document with enough text to assess, at any level from undergraduate coursework to doctoral work and journal manuscripts.
Related pages: Dissertation and thesis help, literature review writing, research proposal writing, capstone project help AI and plagiarism checking services and DrillBit plagiarism report services.
Report only, or report plus reduction
Report only
You want to see where you stand before you submit.
- Both reports as PDFs plus the digital receipt
- Plain English summary of every match that matters
- Advice on what your institution is likely to query
- Turnaround of 30 to 45 minutes on most documents
Report plus similarity reduction
The score is too high and you need it brought down properly.
- Everything in the report only option
- A human editor rewrites the passages that are too close to source
- Citations corrected and quotations formatted to your referencing style
- Over quoted sections rebalanced with your own analysis
- Document re-checked and the new report sent to you
What reduction actually means here
- Genuine paraphrasing by a subject editor who has read the source
- Adding the citations that were missing
- Formatting quotations correctly so they are filtered properly
- Restructuring paragraphs that lean on a single source
- Replacing long block quotes with your own argument
- Identifying self matches so they can be excluded rather than rewritten
The line we draw
- No hidden or white coloured text
- No character substitution from other alphabets
- No invisible spacing tricks to break up matches
- No word spinners or synonym swapping that leaves the meaning mangled
- No AI rewriting tools, on any part of your document
- No claims about a guaranteed percentage before we have seen the file
How it works
Four steps, and most documents come back in 30 to 45 minutes.
Send the file
WhatsApp or email the document in Word or PDF. Tell us your referencing style and any percentage threshold your institution has set.
We run the checks
The document goes through the similarity check and, where there is enough continuous prose to qualify, the AI writing check.
You get the reports
Both PDFs and the digital receipt, with a written summary explaining what each match is and whether it matters.
Reduce and re-check
If the score needs to come down, an editor works through the flagged passages, then we run the file again and send the new report.
Who this is for
Students before submission
You want to know what your marker will see, and you would rather find an uncited paragraph now than in a misconduct meeting later.
Research scholars and PhD candidates
Chapter by chapter checking through a long project, where methods sections and literature reviews naturally carry higher overlap.
Authors preparing for journals
Publishers run their own similarity checks. Clearing the manuscript before submission avoids a desk rejection that costs months.
Students facing an AI allegation
Your work is your own and the detector disagrees. We help you read the report properly and put your evidence in order.
Why students send their files to us
Since 2019
Swift Academic Services has supported students across more than 40 countries, with over 4,800 projects delivered.
Explained, not just emailed
Anyone can forward a PDF. We tell you what the number means for your document and what to do next.
Human editors only
We do not use AI writing tools on any part of your work. Checking a document with software is fine, because it only reads. Writing is never handed to a machine.
Every referencing style
APA, Harvard, MLA, IEEE, Chicago, Vancouver, OSCOLA and more, so citation fixes match what your course actually requires.
Straight answers
If your report is fine, we say so rather than selling you an edit you do not need.
Open when you need us
Reachable daily from 6:00 AM to 12:30 AM IST on WhatsApp, which covers most deadlines in most time zones.
Frequently asked questions
What is the difference between the similarity report and the AI report?
The similarity report measures matched text: how much of your writing overlaps with sources Turnitin can see, and where each match came from. The AI writing report estimates how much of your long form prose was likely generated by AI. They are produced from the same submission but measure completely different things, and a document can score low on one and high on the other.
What is a good Turnitin similarity score?
Most institutions are comfortable below 15 to 20 percent, and many treat under 10 percent as unremarkable. There is no universal pass mark, so the only figure that matters is the one in your own course handbook. Methods chapters and literature reviews often sit higher than the rest of a document for legitimate reasons.
Does a high similarity score mean I have plagiarised?
No. Similarity counts matched text, including quotations you attributed correctly, reference lists and standard subject terminology. A document can score 30 percent and be entirely honest, or score 8 percent and contain one uncited paragraph that causes a problem. The match groups, not the headline number, tell you which situation you are in.
Will my document be stored in the Turnitin database?
Tell us before you send the file if this matters to you, and we will confirm exactly how your submission is handled before we run anything. This is worth asking about, because if a document is added to the standard repository, a later submission of the same work by your university can match against it at a very high percentage.
How long does it take?
Most documents come back in 30 to 45 minutes. Long theses and manuscripts submitted at peak times can take longer, and Turnitin itself queues repeat submissions of the same file, so a re-check after edits is not always instant. If you are against a deadline, say so when you send the file.
Which file formats do you accept?
Word and PDF cover almost everything. Scanned images and photographs of pages cannot be assessed properly, because there is no selectable text for the system to compare, so send the original document wherever you can.
Can you check a full dissertation or thesis?
Yes, and long documents are a large part of what we do. For a full thesis we normally recommend checking chapter by chapter during writing rather than once at the end, because fixing a citation problem in chapter two is far easier before chapter six depends on it.
Why was my own writing flagged as AI?
AI detection scores patterns, not authorship. Heavily grammar corrected text, writing translated from another language, rigid departmental templates and very uniform sentence lengths can all push the figure up. Turnitin prints its own disclaimer on the report stating that the assessment may not always be accurate and should not be the sole basis for action against a student.
Can you reduce my similarity score?
Yes, by fixing the underlying cause. A human editor rewrites passages that sit too close to their source, adds the citations that were missing, formats quotations so they are filtered correctly, and rebalances sections that lean too heavily on one reference. We do not use spinners, character tricks or AI rewriting tools.
Do you guarantee a particular percentage?
No, and be careful with anyone who does. The final figure depends on your subject, your referencing style, how much you legitimately quote and what is already in the database. What we will commit to is that every uncited match is dealt with and that you understand every number on your report before you submit.
Do you use AI to write or rewrite my work?
No. Swift Academic Services does not use AI writing tools on any part of a client document. Running a file through a checker is different, because a checker only reads the text. Writing and rewriting are done by human editors.
Can I get a report for work I have already submitted?
Yes. People often do this when a marker has raised a query and they want to see the report for themselves before replying. We will run the check and walk you through what it actually shows.
What does the digital receipt prove?
It records the submission ID, the date and time, the file name, the page count and the word count for the document that was checked. It is the evidence that a specific version of a specific file was assessed at a specific moment.
Are you affiliated with Turnitin?
No. Swift Academic Services is an independent academic consultancy. We are not affiliated with, endorsed by or a reseller for Turnitin LLC, and we do not sell Turnitin accounts or licences. We run checks for our clients and explain the results.
Send the file, get your report in 30 to 45 minutes
Share your document on WhatsApp or by email, tell us your referencing style and any threshold your institution has set, and we will come back with both reports and a clear explanation of what they mean.