Plagiarism checker for students.
Self-check your essay, paper, thesis, or dissertation before you submit. Catches paraphrasing today, with AI-generated-text detection coming soon — the two signals teachers actually look at. Free up to 2,500 words. Your check stays yours — never visible to your school.
From a 500-word essay to a 60,000-word dissertation.
Different assignments, different concerns. We've laid out the self-check workflow for each — what to check, what to ignore, what to fix before you submit.
Short essay (500–2,500 words)
One sitting, one paste. The Free tier fits a typical class-essay length. Focus on phrasing — uncited paraphrases are the #1 thing flagged at this length.
- Free tier covers a typical essay
- Mobile-friendly — check on your phone
- AI-help disclosure if you used a chatbot
Coursework paper (3,000–8,000 words)
Longer texts get chunked and checked in one batch on Pro. The citation classifier separates your direct quotes from your unattributed paraphrases.
- Pro tier free for 30 days
- Citation-aware report (quote/paraphrase/cite)
- Export PDF for hand-in
Master's thesis (15,000–40,000 words)
Break it into chapters or paste the whole thing — the engine handles either. Re-run after each revision; history saves locally so you can compare versions.
- Chapter-by-chapter check supported
- Version history across edits
- Bibliography auto-excluded from score
Doctoral dissertation (60,000+ words)
Run before submission, run before the defense, run again before the bound copy. The engine flags reused phrases across your own chapters (legitimate at this length) so you can decide.
- Self-citation detection
- Cross-chapter phrase-reuse flag
- EU residency option for your data
Three steps. Twenty minutes. Done before you turn it in.
Self-check is not the same as the school's check. Use this to see what they'll see — and to fix it first.
01 · Draft until you're done
Write the way you actually write. Don't worry about the checker yet. Properly cited quotes, paraphrases with footnotes, your own arguments — write it the right way, then verify.
02 · Run the self-check
Paste or upload. The report classifies every match: Direct quote with citation (good), Paraphrase with citation (good), Common knowledge (excluded), Uncited paraphrase (fix). AI-likely-passage flagging is coming soon (disclose or rewrite). Free tier; no signup needed.
03 · Fix what needs fixing
An uncited paraphrase? Add the citation (or rewrite to be more clearly your own argument). Once AI-suspect-passage detection rolls out, if you used an AI assistant, disclose it where your school's policy says to. Re-run if you made big edits.
Six citation styles. Auto-detected from your bibliography.
Drop your paper in; the engine identifies which citation style you're using and grades each in-text reference. Wrong style for the journal? It flags that too.
MLA · 9th edition
Modern Language Association. Standard for English, literature, humanities. Parenthetical author-page citations + Works Cited. Generator handles books, journals, websites, YouTube videos, podcasts.
APA · 7th edition
American Psychological Association. Standard for psychology, sociology, education, business. Author-date in-text + References list. Handles DOI auto-resolution.
Chicago / Turabian
Notes-bibliography (history, art) and author-date (sciences). Footnote insertion supported on Pro. Common in humanities theses where MLA isn't required.
Harvard
UK + Australia + Commonwealth default. Author-date in-text + Reference list. Subtle differences from APA — the tool catches when students mix them by mistake.
Vancouver
Medical / biomedical. Numbered citations in superscript, References list ordered by citation appearance. PubMed ID resolution supported.
IEEE
Engineering and computer science. Numbered in square brackets, References list ordered by appearance. Auto-formats from CrossRef DOIs and arXiv IDs.
Check from your phone the night before it's due.
Same engine, mobile-aware layout. Paste the doc from Google Drive or Apple Notes; read the report in the share sheet; export PDF and AirDrop it to your laptop.
Share-sheet integration
Send the PDF straight to AirDrop, Mail, Drive, Files, Notion, or your laptop in two taps. No re-export.
Local-first session
If you're not signed in, the session is local — close the tab and nothing's left behind. Sign in to sync across phone + laptop.
30-second checks for short essays
2,500-word essay finishes before you put the phone down. Result page is sortable, swipeable, and renders the highlighted text inline.
Works on slow connections
Submission is async — fire-and-forget. Lose Wi-Fi mid-bus-ride? The report appears when you reconnect. No restart, no re-paste.
Three things you can do that black-box checkers won't let you.
Not just “63% similar” — which intervals?
A score is a number with no provenance. Ask your instructor (or integrity officer) to show you the specific matched intervals from the report, not just the headline percentage. The Noplag report ships every interval with the source URL, the matched characters, and the original snapshot — auditable.
AI detectors over-flag non-native English writers
Liang et al. 2023 measured a 61.3% false-positive rate against ESL writers on standard AI detectors. The Noplag AI verdict is coming soon and will ship with an ESL-adjusted band when the lexical signature suggests non-native English — but if your school used a different tool, that bias is documented science. Cite the paper.
Your school's tool is open-source? You can read it.
If your school checks with Noplag, the engine is Apache 2.0 — github.com/NoplagLabs/noplag-engine. The threshold for “plagiarism vs paraphrase”, the AI verdict calibration, the dedupe logic — it's all in the repo. Read the commit your check ran on (in the X-Engine-Commit header) and disagree with the math in a public issue if you find one.
We don't share your self-check with your school. Ever.
When a student runs a self-check on noplag.com, the result is stored in the student's own account — encrypted at rest, accessible only with the student's passphrase. We don't have an institutional dashboard that surfaces self-checks to instructors. Your school can't search for your name and see what you submitted to verify before turn-in. The only checks an institution sees are the ones the institution itself initiated through its LMS-integration channel — those are flagged in the dashboard, you know when one is happening, and your private self-checks remain entirely separate. We also don't sell, share, or use student documents to train AI models. The Noplag Database (the cross-customer index) adds submitted documents by default, you can opt out (turn it off) anytime, and it excludes K-12 deployments by policy.
Read the privacy policyThe questions students actually ask.
- Will my school see that I ran this self-check?
- No. Self-checks live in your own account, encrypted at rest. We don't have an institutional dashboard that surfaces student self-checks to instructors — the only checks a school sees are the ones it initiated through its LMS integration, and you know when those are happening. Run a self-check the night before you submit. Nobody at your school can search for it.
- How is this different from running my paper through Turnitin?
- Turnitin is an institutional product — the school decides when it runs, what corpus it checks against, and who sees the result. Your school's version flags self-submitted papers if your institution has “Originality Checks for Students” enabled. Noplag's self-check is the opposite: you decide when, you see the result first, and you choose what to fix before the institutional check ever runs.
- I'm ESL — will the AI detector unfairly flag me?
- The Liang et al. 2023 paper measured a 61.3% false-positive rate for ESL writers on standard AI detectors. Noplag's AI verdict is coming soon and will ship with an ESL-adjusted calibration that recalibrates against an ESL-validation set when the lexical signature suggests non-native English. Targeted residual error rate ~3% (documented, not zero). You'll see ESL ADJ next to your verdict once the detector launches and the recalibration kicks in.
- I used ChatGPT for brainstorming — am I in trouble?
- Depends on your school's policy. The AI detector (coming soon) will report a likelihood score — it won't make the call about whether using AI was allowed. Check your assignment instructions and your school's academic-integrity policy. Once it ships, if AI assistance is allowed with disclosure, the report will tell you which sections to disclose; if it's not allowed, it will tell you which sections to rewrite as your own work.
- How free is the free tier really? What's the catch?
- 2,500 words per check, unlimited checks, no signup. No card. No daily cap. The catch: longer documents (theses, dissertations) need Pro ($23/mo, or Premium at $79/mo for full LMS integrations) — but a typical class essay fits in Free comfortably. Student email gets you 30 days of Pro for free.
- Do you save my paper somewhere where another student could find it?
- By default, submitted documents are added to the cross-customer index (the Noplag Database) — but you can opt out (turn it off) anytime, and opted-out submissions live only in your own encrypted folder. Right-to-erasure: delete your account, every fingerprint we hold of you is deleted. We don't sell, share, or use student documents to train AI.
- What if my dissertation is 60,000 words — does it really handle that?
- Yes, on Pro. Submit it as one document or chapter-by-chapter (your choice). The engine chunks it internally; the report stitches back. Version history saves locally, so you can compare a check from January against the same document after May's revisions. EU residency option available if your university's data policy requires it.
- Will it catch a paraphrased Wikipedia passage?
- Partly today. Light rewording that leaves five-word runs intact is caught by L1 winnowing; thorough paraphrasing is what the L2 (MinHash LSH, v1.1) and L3 (vector embeddings, v1.2) layers are being built for, and they have not shipped yet. The report classifies it: Paraphrase (uncited) → flag, Paraphrase (cited) → expected, Common knowledge → excluded. If your Wikipedia paraphrase has a citation, it doesn't count against you.
- What's the citation generator like?
- Pro tier. Pick a style (MLA, APA, Chicago, Harvard, Vancouver, IEEE), drop a DOI / ISBN / URL / arXiv ID into the bibliography panel, and the entry generates. The engine then grades each in-text citation against the bibliography — wrong style, missing pages, formatting drift. Free tier only does the matching; Pro adds the generator.
- What if I disagree with the score?
- Open a public GitHub issue at github.com/NoplagLabs/noplag-engine. The engine is Apache 2.0, the algorithm is in the open, and we treat bug reports from students the same as from anyone. If the score is wrong because of an algorithm bug, it gets fixed and the fix ships to your school's deployment within the next release.
Check it before you turn it in.
Free covers a typical essay. No card, no signup, no “your school will see this” dialog. Your check stays yours.