Skip to content
INDXR.AI
Workflows

YouTube video summarizer: turn a long video into chapter notes

IE
INDXR.AI Editorial
Published August 24, 2026 · Updated August 24, 2026

You want a summary of a video. Maybe a lecture you need to revise from, an interview you want the substance of, a course you are working through, or something you came across and do not have an hour for. INDXR reads the whole transcript, splits it into chapters at the points where the subject changes, and writes worked-out notes under each one, with a timestamp that jumps the player to that moment. It costs 3 credits for a video up to 30 minutes and 1 more for each additional 20 minutes, and a free account includes 25 credits.

What a YouTube video summary looks like

What you get back is not a paragraph but a set of notes. At the top is an overview of a few paragraphs that covers the whole video. Below it are the chapters, each with a heading, a timestamp and the content written out: the arguments, the examples, the figures and the names that came up in that part, rather than a summary of them. More content produces more chapters, and more chapters produce a longer whole, so a long video really does give you a long set of notes. That makes it suited to material you have to do something with, and less to the case where you only want to know what something is about.

A summary of a video: a few-paragraph overview at the top, then the first chapter below it with a heading and a clickable timestamp.
A summary opens with an overview of the whole video, then the chapters beneath it, each with a heading and a timestamp.

A real example

Take a 55-minute lecture on justice, with almost seven thousand words of transcript. The summary came back as a three-paragraph overview plus two chapters, a little over seventeen hundred words together: roughly one word for every four spoken. Two chapters and not ten, because the lecture really is in two parts, and that division comes from the material rather than from the clock. The first chapter covers the trolley problem and its variants. The second begins at 24:40 with Bentham's utilitarianism and the case of Dudley and Stephens. It took about 35 seconds and cost 5 credits.

One chapter of the summary: a heading, a clickable timestamp, and the worked-out notes written out beneath it in full.
Each chapter is written out in full: a heading, the timestamp it starts at, and the arguments and examples from that part of the video.

Timestamps you can check

A summary you cannot check is a summary you have to trust. Every chapter carries the moment it starts, and clicking it takes the player there. That changes what the summary is for: it does not replace the video, it gives you a way into it. Read the chapter list, find the part that matters, jump to it, and watch three minutes instead of two hours.

What it costs

A summary costs 3 credits for a video up to 30 minutes, and 1 more credit for each additional 20 minutes.

Video lengthCredits
Up to 30 minutes3 credits
1 hour5 credits
2 hours8 credits
4 hours14 credits

The price scales because the work scales: a longer video means more chapters and more text to write. A flat price would mean short videos paying for the long ones. There is no subscription, and credits never expire.

How to summarize a YouTube video

Transcribe the video first, with YouTube's own captions, which are free, or with AI transcription at 1 credit per minute if it has no captions. Open the transcript in your library and choose Summarise. It runs in the background, so you can close the tab, and an hour-long video is usually ready within a minute. The summary works on any transcript in your library, whatever its source: YouTube captions, AI transcription, or a file you uploaded yourself.

What it does not do

The summary works from the transcript, so it knows nothing about what happens on screen. Slides, diagrams and demonstrations only reach it through what was said about them, not through what was shown.

And you do not edit the summary in place. It is read-only: to change it you regenerate it, which replaces the current one with a fresh version, or you copy it out and edit that copy wherever you keep your notes. The transcript underneath is editable, and editing it marks the summary as out of date so you can regenerate it, but the summary text itself is not something you rewrite in the library.

Try it

A free account includes 25 credits, which covers pulling captions from a video and summarising it several times over. No subscription and no card.

Frequently Asked Questions

Is there a maximum video length for a summary?
There is no hard limit on the summary itself. If a video has no captions you transcribe it first, and AI transcription is capped at ten hours per file; past roughly five hours the summary keeps its chapters longer rather than adding more of them.
Can I export the summary, or only read it?
You can copy it or download it as a .txt file from the summary view. The overview, the chapter headings, their timestamps and the notes all come across as plain text.
What writes the summary, and where does my text go?
Your summary is written by Gemini 2.5 Flash via the AssemblyAI EU LLM Gateway, so the transcript text stays inside the EU while it is summarised. It runs in two passes: one to find the chapters, and one to write the notes under each.
Where is the summary kept after it is made?
With its transcript, in your library, so it is there whenever you reopen that video. You can regenerate it later, which replaces it with a fresh one.
How many chapters will a summary have?
As many as the material supports. As a rough guide it aims for about one chapter per eight minutes of video, but it follows the actual structure, so a two-part talk can come back as two chapters and a sprawling one as many more.
Will regenerating give me exactly the same summary?
Not exactly. The model is not deterministic, so regenerating can divide the chapters a little differently or word the notes differently, though the substance stays the same.

See also