The whole live web
We do not search a database we built ourselves. We search Google’s index — the largest and freshest there is, updated while you read. A copy posted online yesterday is in there.
Fresh today, not last year
The AI picks the passages worth investigating. Then we search them verbatim on Google — the web as it is today, not a database from last year. Essays, articles, manuscripts, reports.
No credit card. Your document is never stored. Server in the EU.
plagcheck / report
Essay on civic education.pdf
"He regarded popular education as the foundation of a living democracy, in which the individual citizen became ..."
Jensen, K. (2019). Education and the Public, p. 42.
Found on Google · verbatim match
The whole web
is what we compare against
Every flagged passage is searched verbatim in the same Google index you search yourself — updated today. What we will not find is anything behind a login or only in print. We say so rather than pretend otherwise.
Why plagcheck
Most "AI plagiarism checkers" are exactly that — only AI. They can suggest something resembles something else, but they never go and look for it, so they cannot point to the source. We can.
Features
From where we search to a report you can act on.
We do not search a database we built ourselves. We search Google’s index — the largest and freshest there is, updated while you read. A copy posted online yesterday is in there.
Fresh today, not last year
Our model catches rewrites, translations and AI-generated content, not just word-for-word matches. It knows the difference between a quotation and a copy.
Fewer false positives, fewer loopholes
Detailed reports with suggested citations and concrete source references — not just a red line, but a way forward. Ready to act on.
Insight, not guesswork
How it works
A simple process that works for every kind of text.
Drag a document in. Word, PDF, Google Docs, LaTeX. Every format, no conversion.
≈ 2 sec.
Our model reads the text and flags passages that need a closer look, rewrites included.
≈ 15 sec.
Every flagged passage is compared against the whole indexed web, across languages and sources.
≈ 30 sec.
A report that points to the sources, suggests correct citations and gives you grounds to act.
Instantly
Built in the EU
Questions
We search Google's index verbatim through a search API. That covers the open web as it stands right now: essay banks, blogs, news media, public documents and everything else Google has indexed. We do not have our own database, and we do not claim to. What we will not find: closed archives behind a login, printed books that were never digitised, and other students' unpublished work. If you come via opgaver.com, their own archive is searched as well.
Yes. Our model is trained to recognise statistical patterns typical of LLM-generated text, and we cross-reference against known AI output. We are transparent that these are probability assessments, not proof. That's why we deliver context, not verdicts.
Every kind of text document. Academic essays, dissertations and reports, journalism and features before publication, book manuscripts and chapters, marketing copy and posts, legal documents, and ordinary reports and memos. Word, PDF, Google Docs, LaTeX and plain text.
Because we actually go and search for the text, we can always point to the specific page it sits on, rather than saying "this might resemble something". You get a link, not a suspicion. Our report is built to give you grounds to act — not to trigger an automatic verdict.
Our server is in the EU (Frankfurt), and your document is never stored — it exists only in memory while the analysis runs. But the analysis cannot be done without outside help: an excerpt goes to OpenAI (USA) to select passages and write the report, and the selected sentences are looked up on Google via Serper (USA). We do not train models on your texts, and OpenAI states it does not use API data for training. We are ready to sign a data processing agreement. Every detail is in the privacy policy.
Your first check is completely free, no credit card needed. After that we offer flexible plans for individuals and organisations. Contact us for tailored pricing.
Try for free
No credit card, no commitment, no trial clock ticking. Upload a document and see for yourself what is actually out there.