Turn any article into a narrated video lecture

Paste a URL or the full text of an article into Scholarly and the AI reads the entire piece before it writes anything. It identifies the thesis, the supporting evidence, the counter-points, and the conclusion, then drafts a narrated script that mirrors the structure of the original without the filler. Long-form journalism, peer-reviewed papers, opinion pieces, and tutorials all get the same treatment: the AI keeps the load-bearing claims and trims the asides.

As the narration plays, Scholarly pins the key claim, the relevant quote, or the cited statistic to the screen so you can follow the argument visually. Section headers double as chapters, so a 20-minute read collapses into a 5–8 minute video that you can scrub, slow down, or rewatch. The original URL stays linked under the player so you can jump back to the source when a passage demands a closer read.

Articles assume a quiet hour and full attention. Most of the time, students don't have either. A video lecture works in the gaps — on a walk between classes, on a bus, while cooking — and the AI's narration is paced for listening, not scanning. Because the script is rewritten for the ear rather than the eye, sentence length shortens, jargon gets unpacked, and transitions are spoken out loud instead of buried in paragraph breaks.

Video is also the format your brain remembers best when claims and visuals fire together. A highlighted statistic anchored under a narrated sentence is encoded with both an audio and a visual cue, which makes recall during an essay or exam more reliable than a re-read alone. That dual coding is what turns a passive skim into something you can actually quote three weeks later.

Drop in a URL from almost any public source: news outlets, journals, Substack, Medium, blog posts, Wikipedia articles, knowledge bases, and most university lecture pages. Paywalled articles can be pasted as text if you have access. PDF reprints of articles work too — Scholarly parses the PDF, strips the headers and footers, and treats the body the same way it would a URL. Foreign-language articles are translated before the script is written.

You can also paste multiple articles into a single video. The AI will read all of them, find the shared thesis or the points of disagreement, and produce a single explainer that compares the sources. That makes it the fastest way to brief yourself before writing a paper that has to cite three or four sources side by side.

Once the video is generated, it becomes a study object inside Scholarly like any other piece of content. One click turns the highlighted claims and definitions into spaced-repetition flashcards, another generates a practice quiz, and the AI chat sidebar can answer questions about anything the article touched on. The video, the article, and the cards all share a transcript, so a question typed into chat returns an answer linked back to the exact paragraph and the exact timestamp.

This pipeline — article → video → flashcards → quiz — is how you go from "I should probably read this" to "I can defend this in a discussion" in an afternoon. Each step uses a different learning modality, and the AI keeps the source material consistent across all of them so you never lose the thread of the original argument.