Turnitin’s Plagiarism Tool — and What to Use If You Can’t Buy It

Most people searching for Turnitin’s plagiarism detection tool are not looking for an explainer. They have a document, a deadline, and a reason to check it — and they want to buy the thing. That is where this query dead-ends: Turnitin is sold to institutions, not to people. There is no consumer checkout, no per-document price, and no individual tier at any price. If your university or publisher already licenses it, you use it through them. If they do not, you cannot buy your way in.

This page does three things, in order. It states the licensing position precisely, so you stop searching for a purchase page that does not exist. It explains what Turnitin’s database actually contains, so you can judge whether an alternative would give you a comparable answer or a misleadingly different one. Then it routes you by situation — because for one of the three common situations, the honest answer is that no alternative exists, and for another, the correct tool is not the one we earn a commission on.

Tip: try code CASRAI at checkout for 15% off, if the offer is currently active for this program — codes vary by vendor and aren’t guaranteed.

Can an individual buy Turnitin? The licensing answer, precisely

No. Turnitin’s similarity and grading products are licensed to institutions — schools, universities, and organisations — and Turnitin’s own help centre answers the question “Can I purchase an individual Turnitin license?” directly: individual licences are not sold. Access reaches a student or an instructor only through an institution that already holds a licence, which is why the practical next step for anyone affiliated with a university is the library or the teaching-and-learning helpdesk, not a credit card.

Three consequences follow, and they are the reason this query has such high commercial intent and such a low satisfaction rate:

• Price is not the obstacle. A PI willing to pay $500 for one manuscript check still cannot become a Turnitin customer. This is a channel decision, not a pricing tier.
• “Free Turnitin” services are not Turnitin. Sites offering to run your document “through Turnitin” for a fee are reselling access obtained through an institutional account, which typically violates that account’s terms — and, more importantly for you, may deposit your unpublished manuscript into a repository you did not agree to. Treat any such service as a confidentiality risk to your own unpublished work.
• The nearest legitimate route is a sibling product. Turnitin owns iThenticate, which runs the same text-matching engine and does sell to individuals on a per-document basis. That is covered in detail below, and for a large share of readers arriving on this page it is the correct answer.

Does Turnitin check plagiarism and AI text, or are they separate products?

They are two different detectors that happen to be delivered inside the same institutional platform, and conflating them causes real confusion when people go shopping for a replacement.

Similarity is text-matching. It compares a submission against a reference database and reports the percentage of the text that matches something already in that database, with the matched passages highlighted and linked to their sources. It makes no judgment about intent, attribution, or whether a quotation is properly cited.

AI writing detection is a statistical classifier estimating how much of a passage reads as machine-generated. It is bundled into the institutional Similarity/Feedback Studio licence rather than sold separately, and it has been a genuinely contested feature: several universities have switched it off while keeping text-matching on, citing false-positive risk at scale. We cover that history, the vendor’s accuracy claims, and the independent testing that disputes them in Turnitin AI detection vs standalone AI detectors and in AI detection accuracy in higher education.

Why this matters for your purchase decision: if what you actually need is AI-text screening rather than source matching, the institutional-only barrier is much easier to route around, because standalone AI detectors sell directly to individuals. If what you need is source matching against student coursework, it is much harder — for reasons in the next section.

What Turnitin’s student-paper repository contains — and why nothing else has it

Turnitin’s similarity index is computed against three broad source sets: current and archived web content, licensed periodicals, journals and publications, and — the part that has no substitute — papers previously submitted to Turnitin by other students.

That third corpus is the entire competitive moat. It is built from years of coursework submitted through institutional accounts. It is private, it is not on the open web, and it is not licensed to any competitor. Its practical effect is that Turnitin can catch a student submitting an essay that another student at another institution submitted three years ago — a match that is invisible to every web-crawling checker on the market, because the source document was never published anywhere.

So when a comparison page tells you that some tool is “just like Turnitin,” check whether it is claiming coverage of that repository. It cannot have it. For coursework that must be matched against prior student submissions, there is no alternative to Turnitin, at any price. If that is your use case and your institution does not license Turnitin, the honest advice is to raise it as a procurement request internally, not to buy a per-seat tool and assume it does the same job. A general-purpose checker will return a low similarity score on a recycled student essay and you will read that low score as exoneration when it is simply blindness.

If your use case is different — screening manuscripts, triaging submissions to a small journal, checking your own drafts for unattributed overlap with published literature, or screening for AI-generated text — then a per-seat tool such as Originality.ai covers a meaningful share of what you need, and the rest of this page is about where the boundaries fall.

See what Originality.ai actually scans →

Plagiarism checkers like Turnitin: what an alternative can and cannot replicate

Split the question into coverage and workflow. Coverage is where the substantive differences live.

What a web-indexed checker replicates well

• Open-web and published-web overlap. Text lifted from web pages, open-access articles, preprint servers, institutional repositories, theses posted online, Wikipedia, and content-farm output is well covered by any serious commercial index, including Originality.ai’s.
• Self-overlap with your own published work. If your earlier paper is discoverable online, a web-indexed checker will flag text recycling from it — useful when you are trying to avoid an unintentional self-plagiarism problem before submission rather than after.
• Volume triage. Per-word credit pricing makes it economic to screen a large batch of submissions or drafts, which per-document pricing does not.
• AI-text screening in the same pass. This is the capability iThenticate’s individual tiers explicitly do not include, and it is the clearest reason a reader might legitimately choose Originality.ai over the closer-to-Turnitin option.

What no alternative replicates

• The private student-paper repository. Covered above. Structurally unavailable to competitors.
• Paywalled journal full text at the depth publishers get. Crossref Similarity Check matches against full text deposited by Crossref member publishers — subscription content that is not on the open web and not in a general commercial crawl. For a manuscript that will be screened by a publisher using exactly that corpus, checking it against a web index first tells you less than you think.
• The institutional grading workflow. LMS integration, submission inboxes, rubrics, marking, and the audit trail an academic-integrity panel expects are platform features, not detection features. No per-seat tool substitutes for them.

Turnitin vs iThenticate vs Crossref Similarity Check vs a per-seat checker

Prices below were checked against each vendor’s own pricing page in August 2026 and will drift; re-check before budgeting. Turnitin’s own institutional pricing is quote-based and not published.

For the institutional side of that decision — which of the two Turnitin-owned products a university or publisher should license, and why they are not really rivals — see iThenticate vs Turnitin. For how the three matching engines work under the hood, see Anti-plagiarism software: how Turnitin, iThenticate and Crossref Similarity Check work.

Check Originality.ai pricing and credit maths →

What a similarity score does and does not prove

Every tool on this page returns a percentage, and the percentage is the single most misread number in research integrity. Turnitin’s own guidance is explicit that the Similarity tool does not determine whether plagiarism has occurred; it surfaces matching text for a human to interpret.

What a high score can legitimately mean: correctly quoted and cited material, a standard methods paragraph reused with permission, a reference list, a boilerplate ethics statement, a common phrase in a crowded field. What a low score can legitimately hide: a paraphrased argument with no matching strings, a translated passage, an idea taken without text, or a source that is not in that particular index — including, as above, an unpublished student paper.

There is also no field-wide numeric threshold. No vendor, no publisher body, and no regulatory definition of research misconduct sets a percentage at which text overlap becomes misconduct. Figures around 15% or 20% circulate widely; they are local institutional or journal policies, not a standard. If you are building a screening workflow, write down what a flagged report triggers — a human read, not an automatic finding. Our originality report entry and types of plagiarism guide cover how to interpret matched text against the actual categories of plagiarism a panel would recognise.

Per-seat options for journal editors, PIs and unaffiliated reviewers

Three situations account for most of the traffic to this query, and they route differently.

You are an author with one manuscript to check before submission

Buy the iThenticate single-manuscript package. At $125 for up to 25,000 words as of August 2026 it is not cheap for one document, but it is the closest legitimate match to what a publisher will run against your paper, including subscription full text that a web index does not reach. Paying $12.95/month for a per-seat checker instead, and reading a clean report as clearance, would be false comfort — the corpora are different. This is the case where our affiliate product is not the right answer, and we would rather say so than sell you the wrong tool.

You edit a small society journal with no Crossref Similarity Check subscription

This is the strongest case for a per-seat tool. You are triaging many submissions, most of which will be desk-rejected, and per-document pricing does not survive contact with that volume. Originality.ai’s Pro plan gives 2,000 credits per month at 1 credit per 100 words — 200,000 words of scanning for $12.95/month billed annually — and returns AI-text screening in the same pass, which matters given how much of a modern submission pile warrants that check. Use it as a first-pass triage layer, then run the manuscripts that survive triage through Crossref Similarity Check if your publisher provides it, or ask the author to supply an iThenticate report. Related: our guide on AI paraphrasing tools and plagiarism risk, and AI detectors compared for research-integrity offices.

You are a PI, integrity officer or reviewer at a non-subscribing institution

You are screening other people’s drafts — thesis chapters, grant text, a co-author’s contribution — repeatedly and unpredictably, and you have no institutional tool. Per-seat pricing fits the pattern; per-document pricing does not. Accept the coverage limits explicitly: what you are buying is web-and-published-web overlap plus AI-text screening, not a Turnitin equivalent. Our longer Originality.ai review for research-integrity offices goes through the scan types, the report format, and the false-positive behaviour in more detail than fits here.

Who should not buy Originality.ai

Stated plainly, because a page that only lists strengths is not worth trusting:

• Anyone screening coursework against prior student submissions. Only Turnitin holds that corpus. There is no substitute, and a clean report from any other tool tells you nothing about that risk. Route this to institutional procurement.
• Any author doing final pre-submission screening of a manuscript. iThenticate or Crossref Similarity Check reach paywalled journal full text; Originality.ai does not. Use the tool with the corpus your publisher will use.
• Anyone who needs the report to carry evidentiary weight in a formal proceeding. Academic-integrity panels and journal investigations generally expect reports from the platform named in institutional or editorial policy. A report from a tool outside that policy invites a procedural objection before anyone reads the content.
• Anyone hoping a detector score will settle a dispute. No score from any vendor does that. See the section above.

And one thing this site will not help with at all: we do not publish guidance on evading detection or “humanising” AI-generated text, and we do not link tools sold for that purpose. If that is what brought you here, this is the wrong site.

Bottom line: route yourself in one step

Ask what corpus your answer has to come from. If it has to include other students’ unpublished coursework, only Turnitin will do, and the path is institutional procurement. If it has to include paywalled journal literature because a publisher will screen against exactly that, buy the iThenticate individual package or use your publisher’s Crossref Similarity Check. If you are screening the open and published web at volume, on your own budget, and you want AI-text detection in the same scan, a per-seat checker is the right shape of tool and Originality.ai is the one we have researched and recommend for that job.

Start scanning with Originality.ai →

Frequently asked questions

Can I buy Turnitin as an individual student or researcher?

No. Turnitin licenses to institutions only; its help centre states that individual licences are not sold. If you are affiliated with a university, ask the library or teaching-support desk whether your institution already holds a licence. If you are not, the closest legitimate self-serve route is iThenticate, which is Turnitin-owned and sells per-document packages to individuals.

Is there a free plagiarism checker like Turnitin?

Not with comparable coverage. Free checkers match against a limited web index, and none has access to Turnitin’s private student-paper repository or to paywalled journal full text. Free tools are useful for catching obvious unattributed copying from public web sources and nothing beyond that. Be cautious about uploading unpublished work to free services — check what the terms say about retaining your document.

What is the difference between Turnitin and iThenticate?

They run the same text-matching engine and Turnitin owns iThenticate. Turnitin is built around coursework and institutional grading workflows and includes the student-paper repository; iThenticate is built around manuscript screening for publishers, editors and authors, adds Crossref member publisher full text, and sells individual per-document packages. See our full iThenticate vs Turnitin comparison.

Does Originality.ai check the same sources as Turnitin?

No, and the difference is material. Originality.ai matches against a web and published-web index. It does not include Turnitin’s private repository of previously submitted student papers, and it does not reach paywalled journal full text the way Crossref Similarity Check does. It does include AI-text detection in the same scan, which iThenticate’s individual packages explicitly do not.

How much does Originality.ai cost?

As of August 2026, its own pricing page lists a Pro plan at $14.95/month, or $12.95/month billed annually ($155.40/year), including 2,000 credits per month, where one credit covers 100 words — roughly 200,000 words of scanning per month. An Enterprise plan is listed at $179/month, or $136.58/month billed annually, with 15,000 credits per month. Pay-as-you-go credits are also sold as a one-off purchase. Verify current figures on the vendor’s page before committing.

What similarity percentage counts as plagiarism?

There is no field-wide answer. No vendor, publisher body or regulatory definition of research misconduct sets a numeric threshold. Percentages in circulation are local institutional or journal policies. A similarity score identifies text worth a human reading; it does not establish that plagiarism occurred, a point Turnitin’s own guidance makes explicitly.

Can I use a plagiarism checker’s report as evidence in an academic-integrity case?

Only if the tool is the one named in your institution’s or journal’s policy. Reports from tools outside that policy are routinely challenged on procedure before their content is discussed. Where a formal proceeding is in prospect, use the platform the policy names and treat any other tool’s output as an internal signal that prompted a human review, not as the finding itself.