Blog · Updated 2026-05-21
How to Get a Transcript From Any YouTube Video (5 Free Methods, 2026 Update)
Whether you need a transcript for note-taking, blog repurposing, language learning, or academic citation, you have five free options in 2026. We tested all of them on the same 30-minute interview, and they're not equally good. Here's the head-to-head.
TL;DR
- Fastest: Paste the URL into a transcript site like ScribeTube — 4 seconds, no setup.
- Most accurate: The creator-uploaded transcript (if the channel uploaded one). YouTube's built-in panel surfaces it.
- Most flexible: The ScribeTube API if you're a developer, or
yt-dlpif you're CLI-comfortable.
Method 1 — YouTube's built-in transcript panel
YouTube has shipped a native transcript panel since 2008. Most users never see it.
- Open the video in a desktop browser (this feature is hidden on mobile web).
- Click the three-dot menu under the video.
- Click Show transcript. A panel opens on the right.
- Click and drag to select, then copy.
Pros: Official, no third-party tools, exact source of truth.
Cons: Timestamps are embedded in the copied text and clutter it. You can't easily strip them. Not available on mobile web. No translation. No download.
Method 2 — Paste the URL into a transcript website
Sites like ScribeTube, YouTubeToTranscript, and Tactiq let you paste a URL and read the transcript in a cleaner UI. They pull the same caption track YouTube uses, but format it for reading.
The best ones add:
- Clickable timestamps that jump the embedded video to that moment
- Search within the transcript
- Original-language extraction with optional translation
- Copy/download as TXT, SRT, VTT, or PDF
Pros: Fastest method (4–10 seconds). No setup. Works on mobile.
Cons: Some sites paywall basic features or require signup. Pick one that's free for the use case you actually need — for casual extraction, the free tier on ScribeTube (3/day, <20 min videos) is enough.
Method 3 — Browser extension
Extensions like Glasp, Tactiq, and YouTube Summary with ChatGPT inject a transcript panel directly into YouTube's UI. You click an icon → the transcript appears in a side panel without leaving YouTube.
Pros: Zero context switch. Great for marathon watchers.
Cons: Heavier than a website. Most require Chrome (no Safari). Some send video metadata to a remote AI service even when you don't ask for a summary.
Method 4 — yt-dlp (command line)
If you're comfortable in a terminal, yt-dlp is the most reliable tool that exists. It's actively maintained, free, open-source, and works on every video YouTube can serve.
pip install yt-dlp
yt-dlp --skip-download --write-auto-sub --sub-lang en \
--sub-format vtt \
https://www.youtube.com/watch?v=VIDEO_ID
That gives you a .vtt file. To convert it to plain text, pipe through a small parser (or use ScribeTube, which does both steps).
Pros: Most reliable. Works for entire channels and playlists in batch. Zero rate limit.
Cons: CLI required. VTT output needs an extra parsing step for "just the text."
Method 5 — YouTube Transcript API (for developers)
If you're building an app, integrating an HTTP API is cleaner than shelling out to yt-dlp or scraping. Both ScribeTube API and jdepoix/youtube-transcript-api work this way.
curl -X GET "https://api.scribetube.app/v1/transcript?id=VIDEO_ID&lang=en" \
-H "Authorization: Bearer YOUR_API_KEY"
The free tier gives 1,000 requests/month — enough for a side project or MVP. Paid plans start at $19/mo for 25k requests.
Pros: Production-grade. Handles YouTube IP blocks (residential proxy infrastructure). Returns clean JSON. Caches results so repeat requests are sub-300ms.
Cons: Requires an API key. Costs money past the free tier.
Which method should you use?
| Use case | Best method |
|---|---|
| One-off transcript for personal use | ScribeTube website (paste URL) |
| Note-taking from many videos a day | Browser extension or ScribeTube Premium |
| Bulk extraction (channels, playlists) | yt-dlp command line |
| Building an app or product | ScribeTube API |
| Verifying a quote for citation | YouTube's built-in panel |
What about accuracy?
All five methods return the same underlying captions — they pull from YouTube's own caption tracks. If the creator uploaded captions manually, those are near-perfect. If only auto-generated captions exist, accuracy is usually 90–95%, dropping on heavy accents, fast speech, jargon, and music.
If you need broadcast-grade accuracy on a video without good auto-captions, you'll have to run the audio through a paid speech-to-text service (Whisper, Otter, Rev, Sonix) — those are different tools, costing $0.30–$1.50 per hour of audio.
Try it now
The fastest way to confirm a transcript exists for your video is to paste the URL into the box on our home page. If captions exist, you'll see the transcript in 4 seconds. If not, you'll get a clean error rather than wondering.
Last updated 2026-05-21. We update this guide whenever YouTube changes how captions are served (it happens about twice a year).