How to summarize a YouTube video and listen to the gist
Last updated: 2026-10-29
To summarize a YouTube video, you can read its transcript, paste the transcript into an AI chat, or use a tool that does both for you. On a Mac with Corvi, open the video and tap the Fn key: Corvi takes the video's transcript from YouTube and tells you what the speaker said in a few sentences, aloud. Double-tap to ask about a detail.
Why summarize a video at all
Lectures, conference talks, product demos and podcasts are increasingly published as long videos. Many are 40 minutes or more, while the part you need may be three sentences long. Watching at double speed still costs twenty minutes; skimming the timeline is guesswork. A summary tells you first whether the video is worth your time, and if it is, which part to watch.
Option 1: read the transcript yourself
Most YouTube videos have captions, either written by the creator or generated automatically. Under the video, open the description, choose "Show transcript", and the full text appears with timestamps. You can search it with Command-F. This is free and exact, but reading a 40-minute transcript still takes a while, and auto-generated captions have no punctuation.
Option 2: paste the transcript into an AI chat
Copy the transcript into ChatGPT, Claude or Gemini and ask for a summary. This works, but it takes several steps every time: open the transcript, select it, copy, switch apps, paste, write a prompt, read the answer. For one video a week that's fine; for several a day it adds up.
Option 3: one key with Corvi
- Open the video on YouTube in Safari, Chrome, Arc or another browser.
- Tap Fn once. Corvi takes the transcript directly from YouTube and, within seconds, tells you what the speaker said: the main points, in a few sentences.
- Double-tap Fn and ask: "What did she say about pricing?", "Which tools were mentioned?", "Is chapter three relevant to me?". Each answer goes deeper. If the video doesn't contain the answer, Corvi searches the web and names its source.
If a video has no captions, Corvi tries to have an AI model watch it instead, which takes longer.
The summary comes in your language, whatever the video's language: an English talk can be summarized aloud in Spanish, Turkish or any of 35 languages. See listen in your own language.
Stop typing. Stop reading. Talk and listen.
Corvi is free during early access. Setup takes a minute; if you don't like it, drag it to the Trash.
Download free for MacWhen a summary is enough, and when it isn't
| Situation | Summary is enough | Watch the original |
|---|---|---|
| Deciding whether to watch | Yes | — |
| A news explainer or a product update | Usually | If you need the visuals |
| A lecture before an exam | To find the important parts | Watch those parts |
| A tutorial with on-screen steps | To know what it covers | Yes, follow along |
| A video someone sent "for later" | Often | Rarely |
A summary is a fast filter, not a replacement for a lecture you'll be tested on. Use it to decide, then go to the right minute.
Tips
- Ask specific follow-ups. "What are the three steps?" gets a better answer than "tell me more".
- Use it on long comment threads too. Point at the comments and tap Fn to hear what viewers are saying.
- Combine with notes. After listening, hold Fn in your notes app and dictate the two things you want to remember.
Frequently asked questions
Does Corvi need a YouTube account or an extension?
No. It reads the video page in your browser; there is no extension to install.
Does it work with videos in other languages?
Yes. Corvi summarizes into your language whatever language the video is in.
Does it download the video?
No. It reads the transcript YouTube already provides. For videos without captions, it asks an AI model to watch the video.
Does it work on other video sites?
Corvi takes transcripts directly from YouTube. On other sites it summarizes the text on the page, such as a description or an article.
Is it free?
Corvi is free during early access.