Pre-scan capstone projects, master's theses, and PhD dissertation chapters before Turnitin, iThenticate, or your supervisor sees them. Sentence-level highlights show which lines read AI, with perplexity and burstiness signals so you can fix the prose instead of guessing. Calibrated for international and ESL writers, careful with lit review and methods register, and private by default. Free to try. No card.
A thesis is the longest single piece of academic prose most writers ever produce, and one of the most templated. Undergraduate capstone, master's thesis, and PhD dissertation work all share the same chapter-by-chapter rhythm and need the same pre-defense scan.
The thesis stack runs from a 30-page undergraduate capstone to an 80-page master's thesis to a 200-page PhD dissertation. Pre-scanning fits every layer because the institutional report at the end is the same Turnitin or iThenticate AI check, regardless of whether you are defending a senior project or a doctoral monograph.
Thirty to fifty pages of structured argument over a single term. Free tier covers single-section scans up to 5,000 characters. Pro at $19.99 a month, or $14.99 a month on yearly, unlocks 10,000 character pastes and unlimited scans for the final fortnight before your supervisor signs off.
Sixty to a hundred pages, denser citation requirements, supervisors who already know what AI-shaped prose looks like. Sentence-level highlights matter here because master's writing rewards specificity, and a single AI-rewritten paragraph in the lit review can be the one your supervisor questions. The 90-day Pro history is the safety net while you iterate.
Chapter drafts that get scanned chapter by chapter as they come together. The 10,000-character cap forces you to scan in sections, which matches how supervisors actually read drafts. PDF export keeps a defensible record of which version of each chapter was scanned and when, useful when an examiner asks about a chapter you handed in three weeks ago.
A thesis is not one document with one AI score. Each chapter has its own register, its own paraphrase density, and its own false-positive risk. Read the score in context of the chapter type rather than chasing a single number across the whole draft.
Lit reviews are citation-dense and paraphrase-heavy. Every paragraph is, by design, a paraphrase of someone else's work in your own words. Expect scores in the 55 to 70 band even when entirely your own writing. Sentence highlights matter more than the overall number, and the right move is usually to defend specific paragraphs rather than chase a higher total.
Methods chapters read identically across thousands of theses in a discipline because they describe a standard procedure. That uniformity is a strength scholarly, and a weakness as far as classifiers are concerned. Scores often land in the 60 to 75 band. The right defense is that this is conventional methods register, not a rewrite that loses procedural precision.
The framing prose around tables and figures is brief and formulaic by convention. Short paragraphs and stock phrasings such as "Table 3 reports the descriptive statistics" trigger easily. Scan the framing prose specifically rather than the tables themselves, and keep an eye on whether one section drifts noticeably from another.
Discussion is where genuine synthesis happens and where AI drafting is most likely to have crept in if you used it. Healthy scores here run 70 plus. If discussion lands lower than your lit review, treat that as a real signal worth investigating before defense rather than a calibration quirk.
Abstracts are usually 200 to 350 words, which is below the chunk size at which any classifier is reliable. Treat the abstract number as advisory only and read the sentence highlights instead. The same goes for acknowledgements, preface, and appendices: scan only if your committee specifically asks.
Verified institutional emails get the .edu discount on Pro automatically at signup. Pro is $19.99 a month standard, $14.99 a month on yearly, and $13.99 a month with institutional verification. Volume pricing for writing centres and graduate schools on request. Full details on the pricing page.
Billed $89.88/year — Save $30
Billed $179.88/year — Save $60
Billed $359.88/year — Save $120
.edu, .ac.uk, .ac.in, .edu.au, .edu.sg, .edu.ph, .edu.vn, and .ac.ke emails get Pro at $13.99/mo. The discount applies automatically at signup. View full pricing →
Supervisors and committee members are not adversaries. They sign off on the work and want you to defend successfully. Bring AI detection into the relationship proactively rather than reactively, so the first time a flag comes up you have already done the audit.
Tell your supervisor you plan to run a pre-submission AI scan on each chapter, that you are aware lit review and methods can produce real false positives, and ask whether the institutional check your university runs is known to be lenient or strict on academic register. Many supervisors have seen the false-positive issue with previous students and will appreciate the heads-up.
When you send a chapter for feedback, attach the TextSight highlight PDF as a supporting document. Your supervisor does not need to act on it; they just need to know you have done the audit. If the institutional check later flags the chapter, your supervisor already has context for the conversation.
If a committee member raises an AI-detection result, treat it as a substantive question. Walk through the specific flagged paragraph, explain the disciplinary register or legitimate paraphrase that triggered the flag, and reference your earlier TextSight report. A high Authenticity Score with sentences explained as standard register beats a one-line institutional summary.
TextSight is not a Turnitin or iThenticate replacement. Both still dominate university integrity workflows worldwide. TextSight is the pre-submission scan that runs while you edit, so you have sentence-level context before your supervisor's report decides anything.
Write in your usual editor: Word, Google Docs, or Overleaf for LaTeX. Using ChatGPT for an outline, a literature review brainstorm, or to break writer's block is the realistic 2026 default. Write the prose itself in your own voice from your own notes and reading.
Open app.textsight.ai, paste the section, and scan. Free tier handles 5,000 characters in one paste. Pro handles 10,000. For chapters past 10,000 characters, split by natural section and scan each one in turn. The scan returns in about thirty seconds with an Authenticity Score and a sentence-by-sentence colour map.
Above 75, submit. Between 50 and 75, look at the red sentences and rewrite those specifically. Below 50, the section needs more substantial work. The point is to fix prose that is genuinely AI-shaped or stock-templated, not to game the score on prose you wrote yourself.
One round of editing usually moves a borderline score by 15 to 25 points. Re-scan, confirm you are in the safe band, then hand the chapter to your supervisor or queue it for the institutional check. A typical chapter section round-trips in about fifteen minutes.
Thesis writers split roughly into LaTeX users in STEM disciplines and Word or Google Docs users in humanities and social sciences. TextSight reads the prose regardless of the source, but copy-paste honesty matters.
Copy the rendered text into TextSight rather than the LaTeX source. The classifier reads the prose, not the markup, and citation commands or equation environments will throw off scores if pasted directly. The cleanest workflow is to compile, copy the body text from the rendered PDF or output, and scan that. For Overleaf users, the rendered preview makes this quick.
Paste the prose, leave the footnote markers in place if you wish. TextSight will treat footnote bodies as part of the section if you paste them too; usually it is cleaner to scan body prose first and footnotes separately if you want to check them.
Select the section, copy, paste into the TextSight scan window. The clipboard transfer strips Docs formatting so the classifier sees clean prose. The 10,000 character limit on Pro and 5,000 on free is the same regardless of source application.
Drag a DOCX, PDF, or TXT into TextSight if you wrote in Word or Docs and want to preserve formatting context. Pro accepts files up to 10,000 characters per scan and returns the same sentence-level result the paste-in workflow does. Useful when chapter formatting matters and you do not want copy-paste to clip footnotes or section breaks.
A single percentage is not a fix path. The TextSight result panel shows which sentences reacted and why, with paragraph-level rollups for longer chapter sections, so you can edit specific lines instead of rewriting the whole chapter.
Every sentence is colour-coded by its own AI-likeness score. Red sentences clustered in one paragraph are a stronger signal than scattered yellows. Scattered yellows in otherwise structured prose often just mean you were taught to write formally. You read the pattern, not just the headline number.
Perplexity is how predictable your word choices are to a language model. Low perplexity reads AI-like. The score is shown per-sentence on Pro, which is the diagnostic context you need to decide whether a flag is real AI residue or just an unusually well-rehearsed literature review intro.
Burstiness is how much your sentence length and structure vary across the section. ChatGPT defaults to uniform medium-length sentences. Real human writing has bursty rhythm: one short sentence, one long, one fragment. Low burstiness across an entire chapter is the classic AI fingerprint and the one supervisors learn to spot first.
Every scan is retrievable for 90 days on Pro. For a writer iterating across a six-month dissertation cycle, that means every clean chapter scan and every revision is on record. PDF export lets you save longer-term archives chapter by chapter. When an examiner asks about a draft from three weeks ago, you have receipts.
Thesis drafts are protected by FERPA in the US, by GDPR in the EU and the UK, by the DPDP Act in India, and by local equivalents elsewhere. TextSight is designed to honour those rules out of the box, not as a paid setting you have to find.
Thesis chapters submitted for scanning are never used to train the classifier or any other model. This is a contract clause, not a configuration toggle. It applies on the free tier the same way it applies on Pro and Business. Your work is not training data.
The free tier needs no email, no account, no identity. For thesis writers worried about privacy or institutional disclosure, this matters. You can scan a draft chapter without TextSight ever knowing who you are or which university you attend.
Scan history is private to your account. We do not share scan data with universities, supervisors, examination boards, Turnitin, iThenticate, or any third party. Your scans are not part of any institutional record, and your supervisor or examiner cannot pull them from us.
Any scan can be deleted from your history. On Pro you can delete individual records. Data retention is bound to your settings, and a standard DPA is available on Business and Enterprise tiers for university writing centres and graduate-school cohorts.
More for thesis writers.
The undergraduate and master's pre-Turnitin workflow with the four-step round-trip and false-positive defence.
For university →Seven-tool ranking with Turnitin correlation and false-positive rates side by side.
See the ranking →The pre-scan workflow that catches Turnitin flags before your supervisor does.
Read the guide →Free, Starter, Pro, Business. Yearly billing saves 25%. Institutional Pro at signup.
See pricing →Free to try. No card. Institutional Pro at $13.99/mo for verified .edu, .ac.uk, .ac.in, and .edu.au emails.