We didn't want to oversell this. Below is what's actually working in the current version, what's in beta, and what we haven't built yet but plan to.
Your text is broken into sentences. For each one, we look up 5-word chunks against Wikipedia and report the closest matching pages with similarity scores. The matches are clickable.
A rough flag for passages that read like LLM output. Treat it as a hint, not proof — we're still tuning it. Premium plan.
When a sentence matches a Wikipedia page, we generate a quick APA-style reference for it. You can copy it straight into your bibliography.
The scan works for English and French (which is what we needed for ICT University). Other languages will run but matches may be weaker because of the Wikipedia source mix.
Logged-in users can save scans to their dashboard, export the data as JSON, or just print the result page to PDF from the browser.
Run multiple files at once — useful for a lecturer with 40 submissions to mark. On the roadmap for the next semester.
The scan runs in your browser. We only keep it if you save the scan to your account.
Every flagged sentence links to the page it matched. If we can't show you a source, we don't flag it.
Student and lecturer roles, with the lecturer seeing all submissions from their class in one place.
Right now we only check Wikipedia. We're working on broader sources, but we won't pretend it scans the whole internet.
Create a free account — takes about thirty seconds — and you can run a scan straight away.
Create a free account