Self-check your undergrad essays, term papers, and lab reports before they land in Canvas or get a Turnitin AI report. Sentence-level highlights show which lines read AI, with perplexity and burstiness signals so you can revise the prose in your own voice instead of guessing. Useful when the deadline is midnight, the dorm is loud, and you want to know how your draft reads before your professor does. Private to your account, never used to train our model. Free to try. No card.
For four-year undergraduates juggling several courses at once, where the AI policy changes from one professor's syllabus to the next and the same week brings two essays and a lab write-up. The realistic move is to draft, self-check before submission, then revise the sentences that read AI.
An undergrad's writing load is wide and uneven. One professor allows AI for brainstorming, the next bans it outright, and a third never says. You write a literature essay, a history response, and a chemistry lab report in the same seven days, each in a different register. A self-check before you submit means you walk into each course knowing how that specific draft reads, instead of finding out from a Turnitin report after the grade is in.
Five to twelve pages of argument with a thesis, evidence, and a conclusion. This is the register most likely to read AI-like when you wrote it yourself, because a well-organised five-paragraph structure overlaps with how models write. Sentence-level highlights show whether a flag is real AI residue or just a tight thesis statement, so you fix the right lines.
Methods, results, and discussion paragraphs reward uniform, procedural sentences, which is exactly the rhythm detectors associate with AI. A self-check here is about telling genre from residue. Scattered yellow highlights across a methods section usually mean the genre showing through, not anything you need to rewrite.
Midterms and finals stack three or four deadlines into one week. The free tier covers casual single-essay checks up to 5,000 characters. When you are submitting daily, Pro removes the per-day limit and raises the paste size to 10,000 characters so each draft gets the same read.
TextSight does not replace your professor's Turnitin report. It is the scan you run on yourself first, so you read your own draft the way the grading tool will, while there is still time to revise.
Write the essay in whatever you use for class, Google Docs, Word, or the Canvas text box. Using AI for an outline or to get unstuck is a course-policy question, not the thing this scan is about. Write the actual prose from your own notes and reading.
Open app.textsight.ai, paste the finished essay, and run the scan. The free tier takes up to 5,000 characters per scan; Pro takes 10,000. You get an Authenticity Score plus a sentence-by-sentence colour map showing where the draft reads AI-like.
Look at the red sentences first, then the yellow clusters. Rewrite those specific lines the way you would actually say them, add a concrete example from the course, and vary your sentence lengths. You are fixing prose that genuinely reads AI-shaped, not chasing a number on writing that is already yours.
Run it once more to confirm the flagged lines moved, then submit the cleaned essay in Canvas or whichever LMS your course uses. A typical undergrad essay takes only a few minutes to round-trip, which is the point during a week with three deadlines.
Pro is $19.99 a month standard, or $14.99 a month on yearly billing. Full details on the pricing page.
Billed $89.88/year — Save $30
Billed $179.88/year — Save $60
Billed $359.88/year — Save $120
Yearly billing saves 25%. View full pricing →
A lone percentage tells you nothing about where to revise. The result panel shows which sentences reacted and why, so you can fix the specific lines instead of rewriting the whole essay the night before it is due.
Every sentence carries its own AI-likeness colour. A block of red sentences inside one body paragraph is a stronger signal than yellows scattered across the whole essay. Scattered yellows in tidy prose usually just mean you write the way your comp class taught you. 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. Seeing it per sentence on Pro is what lets you tell a real AI residue from a thesis statement you rehearsed and polished until it sounded smooth.
Burstiness is how much your sentence length and structure vary. Models default to uniform, medium-length sentences. Real undergrad writing is bursty: a short punchy line, a long one, an occasional fragment. Flat, even rhythm across a whole essay is the pattern detectors react to most.
Two readings in one scan. The Authenticity Score is the inverse-AI reading at the document level. Plagiarism Risk catches copied passages and citation-risky phrasing at the same time. Running both before you submit means one pass tells you about both the AI side and the sourcing side of an essay.
A single number is not proof that anyone used AI. A self-check is for catching how your own draft reads while you can still revise it honestly, and for keeping a record if a flag ever turns out to be wrong.
A low Authenticity Score means your essay reads more AI-like to the classifier. On its own it does not prove you used AI, and it does not mean a professor will accuse you. False positives are real, especially for non-native English writers and for the highly structured prose that comp classes reward, where the phrasing overlaps with AI defaults.
Save the report. It holds the text you scanned, the Authenticity Score, the sentence-level flags, and the timestamp. If an integrity conversation comes up, that gives you something specific to point at, a draft history and a set of flagged lines you already revised, instead of a vague "I really did write it."
Be honest with yourself if the score lands below 50 on prose you thought was mostly your own. AI-assisted drafting tends to drift further into a final essay than students expect. The fix is to rewrite those paragraphs from your own notes and in your own voice, not to run them through another tool.
A detector number is one input, not an academic-integrity ruling. The fair process is sentence-level evidence, a conversation, and a look at earlier drafts before any decision. That cuts both ways: it protects you from a wrongful flag, and it is the reason a self-check plus your own revision history is worth keeping.
Most US campuses run Canvas or Blackboard, often with Turnitin's AI check switched on at submission. Native plugins are not shipped yet, so here is what works today and what is coming.
Write inside Canvas, Blackboard, Brightspace, or Moodle as usual. Before you submit, copy the finished essay into TextSight, revise the flagged lines, and paste the cleaned version back into the submission box. A typical undergrad essay takes only a few minutes to round-trip this way.
If you wrote in Google Docs or Word, drag the DOCX, PDF, or TXT straight into TextSight on Pro, up to 10,000 characters per scan. You get the same sentence-level result without retyping or losing formatting in copy-paste.
The Chrome extension runs a one-click scan from any page, including a Canvas submission view. Handy for checking a paragraph of your own draft, an assignment prompt, or a peer review without switching tabs. Available on Starter and above.
Canvas, Blackboard, Brightspace, Moodle, and Google Classroom plugins are on the roadmap. We are not putting a date on it while each platform keeps changing its plugin rules term to term. A plugin that works is worth more than a wrapper that breaks every semester.
A self-check is only useful if it stays private. Your scans are yours, nothing about them reaches your college, and your text is never fed back into our model.
Essays you submit for scanning are never used to train our model or any other. That holds on the free tier exactly as it holds on Pro and Business. It is a stance, not a setting you have to go and find.
The free tier needs no email and no account. If you are wary about leaving a trail, you can scan an essay without us ever knowing who you are or which college you attend.
Scan history is private to your account. We do not share scan data with colleges, instructors, Turnitin, or any third party. Your scans are not part of any institutional record, and your professor has no way to pull them.
Any saved scan can be deleted from your history, and on Pro you can remove individual records. Our privacy practices are GDPR-aware, and a standard DPA is available on Business and Enterprise for writing centres and tutoring teams.
The full student landing page with the false-positive defence and the academic tone preset.
For students →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 professor does.
Read the guide →Free, Starter, Pro, Business. Yearly billing saves 25%.
See pricing →Free to try. No card. 3 scans a day on the free tier.
How TextSight fits other teams and workflows.