Video to notes
Video to notes: a YouTube video or recording as structured notes
Video to notes conversion is mostly subtraction: paste a YouTube link or upload a recording, and AI drops the asides, repeats and filler, then lays out every topic of the talk as a heading with the facts, numbers and names beneath it, free for any YouTube video with captions. It is one action of the AI video summarizer, it covers the whole recording rather than its opening, and the result saves as PDF or DOCX. Files use minutes.
Updated
On this page
How to turn a YouTube video into notes
To turn a YouTube video into notes takes four steps, and the AI starts only at the third click, not before.
- Paste the link. Put a YouTube address in the field at the top of the page, or drop a recording from your device instead.
- Open Detailed notes. When the transcript appears, press the Detailed notes tile in the Understand group. Nothing is generated before that click.
- Check a point against the video. Every line of the transcript beside the result has a timestamp; click one and the player jumps to that moment.
- Save the result. Copy it into Obsidian, Notion or OneNote, or download it as PDF or DOCX.
The same video to notes converter takes recordings: a class exported from Zoom as MP4, a webinar saved as MOV, a talk recorded on a phone as M4A. Those go through speech recognition before the outline, as described in the guide to speech recognition for video files. A long upload is condensed part by part, so the outline of a 2-hour seminar reach its final question as well as its first slide.
What the notes look like
They come as an outline: one short topic line per part of the lesson, and under it the points in the order the speaker made them.
Illustrative notes for a 40-minute history lesson, not a real recording:
Causes of the revolution
- France was close to bankruptcy after the Seven Years’ War and the American war.
- Louis XVI called the Estates-General in May 1789, the first since 1614.
From the Estates-General to the Bastille
- The Third Estate declared itself the National Assembly on 17 June 1789.
- Crowds in Paris took the Bastille on 14 July 1789.
The Terror
- Louis XVI was executed in January 1793.
- The Terror ended when Robespierre was executed in July 1794.Dates, names and figures survive the cut because the outline is written from the transcript, not from a guess about the topic. If a line looks wrong, the timestamped transcript next to it shows exactly what was said.
Which recordings make good notes
Talks with a clear structure give the cleanest outline: a lesson, a lecture, a conference keynote, an explainer or a long interview.
- Works well: one or two speakers, a topic announced at the start, numbers and names said aloud, as in a university lecture or a product tutorial.
- Works, with gaps: a panel where four people interrupt each other; the outline follows the topics, not who said what.
- Works poorly: vlogs, reaction clips and music, where little is said or what is said has no structure.
- Doesn’t work: silent screen recordings and slideshows over background music, because there is no speech to transcribe.
If you are unsure, start with a summary: five to seven points show in seconds whether the talk has enough substance for a full outline.
Notes, summary or transcript: which one to take
Detailed notes sit between a summary and a transcript: shorter than every word, longer than a handful of points.
| Output | Size for a 30-minute talk | What it keeps |
|---|---|---|
| Summary | 5–7 sentences | the main points only |
| Detailed notes | an outline of every topic | points, facts, figures and names, in order |
| Pro Transcript | the full text in paragraphs | every sentence, minus filler words |
| Raw transcript | every line with a timestamp | exact wording and timing |
Choose the outline when you will come back to the content; choose a summary when you only need to know what a clip says. The raw text is always there underneath, and for a captioned video the full transcript with timestamps is free on its own, without any AI.
Studying from video notes
AI notes work best as a first draft of your own, not the final version.
- Paste the result into your notes app and add your own comments under each topic line.
- Ask a custom prompt for ten revision questions with answers, drawn from the same transcript.
- Keep a recorded class and what you made from it together: the guide to summarizing lecture recordings covers files from Zoom, Panopto exports and phone recorders.
- Read a lecture given in another language in yours by switching the transcript language first.
A YouTube note taker that writes while you watch splits your attention between the player and the page. Here the outline is written after the transcript loads, so you can watch with nothing to type.
What’s free and what uses minutes
Notes from a captioned YouTube video are free with no sign-up; a file or an uncaptioned clip is transcribed first, from $0.033 per minute.
- Captioned YouTube video: an outline, a summary and every other AI action at no cost. On a day of very high demand, free AI may pause until the next day.
- First recording without captions: free with no account, in full up to 10 minutes, the first 10 minutes of a longer one.
- After that: one minute of balance per minute of recording, with every AI action on it included; signing up adds 10 minutes of transcription.
What video to notes AI doesn’t do
Video to notes AI reads speech only, so a formula on a whiteboard or a diagram on a slide appears in the outline only if the speaker reads it out.
- No screenshots or images in the result.
- No timestamp on each bullet; the transcript keeps the timing.
- No flashcards, quiz or mind map yet: those tiles are marked “soon”.
- No links from Vimeo, Coursera or a university portal; download the recording and upload the file.
For everything else the AI can do with the same transcript, from an article to a chat, go back to the overview of AI actions for video.
Frequently asked questions
Is turning a video into notes free?
Yes, for a YouTube video with captions: this action and every other one on it are free, with no sign-up. A file or a video without captions is transcribed first, from $0.033 per minute, and the first such recording is free up to 10 minutes.
Do I need an account?
No for captioned YouTube videos and for your first recording. Whole files after that need sign-in, which works through a link sent to your email.
Can I get notes from a video without captions?
Yes. Whisper speech recognition writes the transcript first, for one minute of balance per minute of video, and this action then costs nothing extra.
Is my recording stored?
No. An uploaded original is deleted from the server once its sound is extracted, and the sound pieces are deleted after recognition. The result stays in your browser, or in your account if you are signed in.
Can it make flashcards or a quiz from a lecture?
Not as a tile yet: flashcards and a quiz are marked “soon”. A custom prompt can already write revision questions with answers as plain text.
What language are the notes in?
In the language of the video. To get English notes from a French lecture, choose English in the language menu before you ask.