How to Show / View a Transcript on YouTube

I open a YouTube video, want to read along instead of scrubbing through it, and the transcript is not under any button I can see at a glance. It is there on most videos, just tucked one layer deeper than people expect. This guide is the exact path I click to show a transcript on YouTube, on both desktop and the mobile app, plus what to do when the option is missing and how to pull the transcript out as text when one video turns into a hundred.

Everything below I clicked through and ran myself in June 2026.

Where is the transcript on YouTube?

The transcript on YouTube lives at the bottom of the video description, behind the ...more button, and it opens as a scrolling panel to the right of the player. It is not on the main control bar with the like and share buttons, which is why it feels hidden. YouTube generates that transcript from the video’s caption track, so any video with captions has one.

There are two ways into the same panel on desktop, and which one you see depends on the layout YouTube serves you:

Both land on the same transcript panel. The description route is the one YouTube’s own help page documents, and it is the more reliable of the two. The three-dot menu route appears on some accounts and video types and is the one people search for as the “youtube show transcript three dots menu” path.

How do you show a transcript on YouTube on desktop?

To show a transcript on YouTube on desktop, click ...more directly under the video title, scroll to the bottom of the expanded description, and click the Show transcript button. The transcript opens in a panel on the right side of the page and scrolls automatically to follow the spoken audio as the video plays.

Here are the steps in order:

1. Open the video in a desktop browser.
2. Under the title and channel name, click ...more to expand the full description.
3. Scroll to the very bottom of the description text.
4. Click Show transcript.
5. The transcript panel opens on the right. Each line carries a timestamp.

Once the panel is open, a few controls are worth knowing, because they answer most of the follow-up questions:

YouTube’s official help page confirms the core behavior: in the description, click Show transcript, and the transcript scrolls with the video so you can click a line to jump to that part, per YouTube Help on viewing transcripts. That same panel also surfaces video chapters and, on some videos, mentioned people and places.

The timestamp toggle matters more than it looks. With timestamps on, every line is prefixed with a time like 12:64, which is useful for navigation and noise for reading. Toggling them off gives you a clean block of text, which is the version most people actually want to copy. That copy step is the next thing readers reach for, so it has its own section below.

How do you show a transcript on the YouTube mobile app?

To show a transcript on the YouTube mobile app, tap the video title to expand the description, scroll down past the stats, and tap Show transcript. A scrollable transcript opens over the lower half of the screen and highlights the current line as the video plays, the same way it does on desktop.

The mobile steps, on both the iOS and Android apps:

1. Open the video in the YouTube app.
2. Tap the video title, or the down arrow to its right, to expand the description.
3. Scroll down to the bottom of the description.
4. Tap Show transcript.
5. Tap any line to jump the video to that timestamp.

Two differences from desktop are worth flagging. The transcript opens as a bottom sheet across the lower screen, because the layout is vertical. And there is no timestamp toggle on mobile, so every line keeps its time prefix. The mobile browser (Safari or Chrome on a phone) behaves like a cramped desktop, so if the button is hard to reach, the app is the smoother route.

The button only shows up under the same condition as desktop: the video needs captions. When it is missing entirely, the cause is almost always the caption track, which is what the next section covers.

Why is there no Show transcript option on a video?

There is no Show transcript option on a video when that video has no caption track, because YouTube builds the transcript directly from captions. No captions means there is nothing to lay out as a transcript, so the button never renders. This is the single most common reason the feature looks broken when it is actually just absent.

Captions on a video come from one of two places, and a video can lack both. One is human transcription, where the uploader (or a paid service) types the text and uploads a subtitle file. The other is YouTube’s own automatic transcription, where speech recognition transcribes the audio for you:

YouTube’s automatic captions cover a wide range of languages, so most spoken-word uploads get a transcript even when the creator adds nothing. The official list spans more than 60 languages, including English, Spanish, Hindi, Arabic, Japanese, and Portuguese, per YouTube Help on automatic captions. Google is candid about the limits: it states the captions “are generated by machine learning algorithms, so the quality of the captions may vary,” and may misread speech “due to mispronunciations, accents, dialects, or background noise.”

That leaves a few specific cases where the button stays gone:

• The creator disabled captions for the channel or that upload. Automatic captions can be turned off, which removes the transcript.
• The video has no speech, such as a music video or ambient footage, so speech recognition produces nothing.
• The video is brand new and automatic captions have not finished processing yet, which can take time after upload.
• Your language has no track. If captions exist only in a language you are not viewing, switch the language dropdown in the panel once it is open.

When a transcript does exist but you want it out of the panel and into a document, the copy step is where most people get stuck next.

How do you copy or save a YouTube transcript?

To copy a YouTube transcript, open the transcript panel on desktop, toggle off the timestamps, drag-select the lines, and copy them with the keyboard. YouTube has no one-click download button for transcripts, so copying the on-screen text is the manual route, and it works fine for a single video.

The clean way to do it:

1. Open the transcript panel (steps above).
2. Click the three-dot icon at the top of the panel and choose Toggle timestamps so only the text remains.
3. Click at the start of the first line and drag to the end of the last line.
4. Copy with Ctrl + C (or Cmd + C on a Mac) and paste into your document.

This is serviceable for one transcript, but it falls apart the moment you need ten: every video means expanding the description, opening the panel, toggling timestamps, and hand-selecting text, with no way to script the clicks. It also gives you raw text with no structure, no per-line timing you can use and no machine-readable format. For a content audit, a dataset, or anything you plan to feed into another tool, you want the transcript as structured data, which means pulling it programmatically. That is the next section.

How do you get YouTube transcripts in bulk or as text?

To get YouTube transcripts in bulk or as clean text, you fetch the caption track programmatically instead of clicking the on-page panel. There are two routes I tested: an open-source Python library that scrapes the same data the web client uses, and a transcript API that returns the parsed lines directly. Both give you the transcript as structured data with per-line timing, which the Show transcript panel will never export.

A quick word on the wider tooling first. Plenty of third-party transcription software and no-code automation flows (a Zapier “new video” trigger feeding a transcription step, for example) exist for this job. They are convenient for a one-off, and most are slow and rate-limited at volume. The two routes below are the ones I keep coming back to, with tips on where each one fits.

Pulling one transcript with Python (the open-source route)

The youtube-transcript-api library fetches a transcript by video ID without a browser or an API key. I installed version 1.2.4 and ran it against a public talk in June 2026. Here is the minimal script and what it actually returned:

# pip install youtube-transcript-api  (tested on v1.2.4)
from youtube_transcript_api import YouTubeTranscriptApi

api = YouTubeTranscriptApi()
fetched = api.fetch("arj7oStGLkU")           # any YouTube video ID

raw = fetched.to_raw_data()                  # list of dicts
print(len(raw), "lines")                     # -> 315 lines
print(raw[0])
# -> {'text': 'So in college,', 'start': 12.645, 'duration': 1.37}

That run returned 315 transcript lines, each a dict with text, start, and duration. Listing the available tracks on the same video showed 50 language versions, each flagged as manually created or auto-generated and whether it is translatable. The library is honest about its foundation: its own docs note it “uses an undocumented part of the YouTube API, which is called by the YouTube web-client. So there is no guarantee that it won’t stop working tomorrow,” per the youtube-transcript-api project page.

That last line is the catch. Because it scrapes the web client’s internal calls, it breaks when YouTube changes them, and it gets your IP blocked at volume. The maintainers ship proxy support precisely because requests from datacenter IPs get throttled once you scale up. For a handful of videos it is great. For a pipeline, you inherit the maintenance.

Pulling transcripts at scale with a transcript API

A transcript API takes the blocking and parsing off your plate: you pass a video URL, it returns the transcript as JSON, with the proxy rotation and the undocumented-endpoint churn handled on the server. I run my pulls against our YouTube transcript endpoint, which follows the same request shape across its video, caption, and transcript routes:

curl "https://api.youtubescraperapi.com/api/v1/youtube/transcript?url=https://www.youtube.com/watch?v=dQw4w9WgXcQ&api_key=$API_KEY"

The pattern is one HTTPS call per video: base https://api.youtubescraperapi.com/api/v1, the youtube site slug, the resource (transcript), the video url, and your api_key. The JSON that comes back carries timestamped segments (each with start, duration, and text), the combined text, and is_generated plus a source of native or asr; a format parameter (both, segments, or text) returns just the timed lines or just the combined string, and units switches the timing between seconds and ms. When a video has no human captions, the endpoint returns YouTube’s auto-generated (ASR) track instead of failing, so most spoken-word videos still resolve, while a truly caption-less video comes back with transcript_available: false and a reason rather than an error. To loop over a channel or a saved list of IDs, you call the same endpoint per URL and collect the JSON, with no panel or IP rotation to manage. Grab a key on the youtubescraperapi.com sign-up page and read the parameters on the YouTube Transcript API reference; for the subtitle files (SRT/VTT) instead of the plain transcript, the YouTube caption scraper covers that shape.

Here is how the three routes compare for the job of getting transcript text out of YouTube:

If you only ever read along with one video at a time, the built-in panel is all you need and this is where you stop. If you are collecting transcripts to summarize, search, or analyze, fetching them as data is the route that holds up. My deeper walkthrough of the extraction code is in how to get the transcript of a YouTube video, and once you have the text, summarizing a YouTube transcript with AI covers what to do with it. For a side-by-side of the tools, see the best YouTube transcript scrapers in 2026.

A quick note on the rules

Reading a transcript through YouTube’s own Show transcript feature is just using the product as built. Collecting transcripts at scale is a different activity, and YouTube’s Terms of Service restrict accessing the service “through the use of any automated means (such as robots, botnets or scrapers)” except through the API or with written permission. I am not your lawyer, and I walk through where that line actually sits for data collection in is scraping YouTube legal. Worth a read before you point a script at thousands of videos.