Read a web page
Turn any article link into a source-grounded breakdown
Paste a link to a long article or web page. Novus Learn fetches it, extracts the real text, and builds the same summary, study notes, flashcards, quiz, visuals, and Ask terminal it builds from Wikipedia — with every claim linked back to the page. Only the URL is sent to our server, to fetch the page; the breakdown is built and stored on your device.
How this works and stays private
- Only the URL is sent to our server, and only to fetch the page — nothing is stored there.
- The breakdown is built in your browser and saved only on this device (IndexedDB) until you delete it.
- Answers use the page’s own sentences — nothing is generated or invented — and every claim links back to the source.
- Private, internal, login-only, or non-article links are refused with a clear message.
What “read a web page” does
Reading to learn and reading to skim are different jobs, but most of the web is built for skimming. Read a web page takes a single article URL and rebuilds it as a study space: a quick-read summary, structured study notes, flashcards, a quiz, a guided lesson, a concept map, and a downloadable source list — all from the page you gave it.
The point is not to summarise and move on. It is to let you actually work through an article the way you would work through a textbook chapter, while keeping every claim tied to the exact sentence it came from. If the page cannot support a point, Novus Learn leaves it out rather than filling the gap with a guess.
How it works
Paste a public article link
Give it the URL of an article, blog post, documentation page, or other public web page. Novus Learn fetches the page once, on the server, only to read it — it stores nothing there.
It extracts the real content
A readability pass strips the navigation, ads, and boilerplate, keeping the article body, headings, and metadata (title, author, publish date, canonical link) so the breakdown stays anchored to the original.
It builds a source-grounded project — in your browser
The study material is assembled on your device and saved locally (IndexedDB). Claims quote the page's own sentences, and each one links back to the block it was drawn from.
It tailors the breakdown to the page type
A how-to, an FAQ, a documentation page, and a news article are not the same, so the tabs and study tools are re-ordered to match what the page actually is — without ever inventing a section the page does not support.
Why it stays grounded in the source
There is no language model writing answers here. Every generated claim is an extracted sentence from the page, and every derived connection is labelled as derived. That means you can trust the breakdown the same way you trust the article — and check it in one click.
It also means the value Novus Learn adds is real editorial work: pulling out the load-bearing claims, connecting concepts, and turning a linear page into flashcards, a quiz, and a map you can study from — not a rewrite of someone else's words.
What you get from one link
- A quick-read summary and a full study-notes breakdown.
- Flashcards and a safe, source-checked quiz.
- A guided lesson that walks the material in order.
- An interactive concept map of the ideas and how they connect.
- A downloadable, fully-referenced source list for citing your work.
- An “Ask this page” terminal that answers only from the page's own sentences.
What it will not do
It will not open private, internal, login-only, or non-article links — those are refused with a clear message. It will not invent facts, ratings, or sections a page does not contain. And it will not send your reading to a third-party tracker: the breakdown lives on your device until you delete it.
Frequently asked questions
Does Read a web page work on any URL?
It works on public article-style pages. Private, internal, login-only, and clearly non-article links are refused with an explanation instead of a broken result.
Is the page content sent anywhere or stored?
Only the URL reaches the server, and only so it can fetch the page to read it — nothing is stored server-side. The finished breakdown is built in your browser and saved on your device (IndexedDB) until you remove it.
Does it use AI to write the summary?
No. The summary and answers are assembled from the page's own sentences. Nothing is generated or paraphrased by a language model, and each claim links back to its source text.
Can I study the result offline later?
Yes. Once a page is built into a project it is stored locally, so you can reopen it, study the flashcards, take the quiz, and read the notes without rebuilding it.
How is this different from pasting the article into a chatbot?
A chatbot generates new text and can invent details. Novus Learn only extracts and organises what is actually on the page, keeps every claim traceable, and turns it into study tools rather than a paragraph of prose.