The short answer: on a computer, open the video, click ...more under the title, scroll to the end of the description and click Show transcript. A panel with the full text opens next to the video. Turn off the timestamps from the panel's three-dot menu, select the text and paste it wherever you need it. If there is no transcript button, or the captions are too poor to use, transcribe the video's audio file instead.
That covers most cases. The rest of this guide explains four ways to get the text of a YouTube video, which one fits your situation, and what to do when the obvious route fails. It covers the transcript panel on desktop and mobile, copying it without timestamps, downloading caption files from your own channel, and transcribing the audio when YouTube has nothing usable.
Which method should you use?
| Method | Works when | What you get | Cost and effort |
|---|---|---|---|
| 1. Transcript panel | The video has captions (uploaded or automatic) | Timestamped lines on screen | Free, under a minute |
| 2. Copy without timestamps | Same as above, on a desktop browser | Plain text in your own document | Free, a few minutes of cleanup |
| 3. Download from YouTube Studio | It is your own video | A caption file (SRT, VTT or SBV) with timings | Free, a few clicks |
| 4. Transcribe the audio file | No captions, poor captions, or an unpublished video you have the file for | Punctuated text with paragraphs; speaker labels on Pro | 15 free minutes a month to test; full text with a trial or paid plan; a few minutes of processing |
Rule of thumb: try methods 1 and 2 first, because they cost nothing. Use method 3 if you own the channel. Use method 4 when the text has to be readable as a document, not just searchable.
Method 1: open YouTube's built-in transcript
Every YouTube video with captions also has a transcript. The transcript is just the captions laid out as a scrolling list instead of being shown one line at a time over the picture.
On a computer
- Open the video on youtube.com in any desktop browser.
- Under the video title, click ...more to expand the description.
- Scroll to the bottom of the description and click Show transcript.
- A Transcript panel opens on the right, next to the video, with one line for every few seconds of speech. Click any line to jump to that moment in the video.
- If the video has captions in more than one language, use the language selector at the bottom of the panel to switch between them.
YouTube moves this button around from time to time. If you do not see it in the description, look for a three-dot menu next to the Share and Save buttons under the player. Older layouts placed Show transcript there.
On a phone or tablet
- Open the video in the YouTube app.
- Tap the title or the description area to expand it.
- Scroll down and tap Show transcript.
The mobile transcript is fine for reading along or finding a moment in the video. It is awkward for copying: selecting text in the app is limited. If you need the text in a document, open the video in a desktop browser.
Method 2: copy the transcript without timestamps
By default every line in the panel starts with a time code such as 2:14. Pasted into a document, that becomes a column of numbers mixed into the text, and cleaning it up by hand takes longer than reading the video. Remove the time codes before you copy:
- Open the transcript panel as described above.
- Click the three-dot menu at the top of the transcript panel and choose Toggle timestamps. The time codes disappear and only the text remains.
- Click just before the first word of the first line. Scroll to the end of the panel, hold Shift and click just after the last word. For a short video you can simply click and drag.
- Copy with Ctrl+C (Windows) or Cmd+C (Mac) and paste into Google Docs, Word or a text editor.
What you have now is the raw caption text. Each caption line becomes a short line of its own, so a quick tidy-up is usually needed:
- Join the lines. Use your editor's find-and-replace to replace line breaks with spaces. Then add paragraph breaks back where the topic changes.
- Check the punctuation. Captions uploaded by the creator are usually punctuated. Automatic captions often have little or none.
- Fix names and terms. Product names, surnames and jargon are where caption errors cluster. The same word is usually misheard the same way every time, so one find-and-replace fixes every instance.
Need timestamps for quoting or citing? Copy the text with them switched on instead. A time code next to a quote lets anyone check it against the video.
Method 3: download the caption file from YouTube Studio (your own videos)
If you uploaded the video, you can download its captions as a file instead of copying them from the page:
- Open YouTube Studio and choose Subtitles in the left-hand menu.
- Click the video you want.
- Find the language row you need. Automatic captions show up as a separate entry marked as automatic.
- Open that row's options menu and choose Download, then pick a format: .srt, .vtt or .sbv.
All three formats are plain text files with a timing line before each block of text. SRT is the most widely supported format for subtitles elsewhere, in video editors or on other platforms. If you only want the words, open the file in a text editor and delete the numbers and timing lines, or paste it into any free "SRT to text" converter.
Studio also lets you download the original video file you uploaded. Keep that in mind for method 4: the original upload has better audio than any copy.
Method 4: transcribe the audio when YouTube's captions are missing or not good enough
Captions are not always there, and when they are they are not always usable. You need a different route when:
- the creator has turned captions off, or YouTube never generated any;
- the automatic captions are an unbroken stream of lowercase words with no punctuation, which happens a lot in long talks, panels and interviews;
- you need to know who said what, and captions do not label speakers;
- the video is yours and not yet published, so there is nothing on YouTube to read.
In all of these cases, transcribe the audio directly. This is how it works on SpeakToWords:
- Get the file. For your own video, use the original export from your editing software or download it from YouTube Studio. Downloading other people's videos is restricted by YouTube's terms, so for someone else's content, ask the creator for the file or stick to methods 1 and 2.
- Create a free account. The free plan includes 15 minutes of transcription every month, with files up to 30 minutes, and no card is required to start.
- Upload the video as it is. MP4, MOV, M4V and 3GP video files are accepted, so you do not need to extract the audio track first. Audio files work too: MP3, WAV, M4A, AAC, OGG, OPUS, FLAC, WMA, AIFF and WEBM. The size limit is 500 MB per file.
- Choose the language. Leave it on automatic detection, or set the transcript language yourself if the video mixes languages.
- Wait a few minutes. Processing runs on the server and the dashboard shows the progress of each job. You do not need to keep the tab open.
- Read and use the transcript. The text comes back punctuated and split into paragraphs. On the free plan you see an opening preview of each transcript. The full text unlocks with a trial or a paid plan, which also let you edit it and download it as a text file. On Pro, speakers are identified and labelled, and you can get an AI summary or a translation.
How long a file you can upload depends on the plan: 30 minutes per file on Free, 2 hours on Basic and 10 hours on Pro. Current prices and monthly minutes are on the pricing section of the homepage.
The same steps work for any recording, not only YouTube videos. The guide to converting MP3 and other audio to text goes into formats and file sizes in more detail.
Accuracy, languages and timestamps
How good are YouTube's automatic captions?
It depends on the audio. One clear speaker talking into a decent microphone usually gets readable captions. Accuracy drops with background music, several people talking over each other, strong accents, technical vocabulary and poor recording quality. The usual weak points are always the same: little or no punctuation, no speaker labels, and names spelled the way they sound.
Captions uploaded by the creator are a different matter. They are typed or edited by a person and are usually the most accurate text you will find. In the transcript panel's language selector, automatic captions are labelled as auto-generated, so you can tell the two apart.
Languages
YouTube generates automatic captions only for a limited set of spoken languages. That list is in YouTube's own help pages, and for other languages there will be no transcript unless the creator uploaded one. Viewers can also use auto-translate (player settings, then Subtitles/CC) to read machine-translated captions in another language. The translation inherits every recognition error in the original captions.
SpeakToWords detects the spoken language automatically. You can also choose the transcript language yourself from 20 options, including English, Spanish, French, German, Portuguese, Italian, Polish, Ukrainian, Russian, Turkish, Arabic, Hindi, Chinese, Japanese and Korean.
Timestamps
The YouTube transcript panel and caption files are built around timestamps, which is what you want for subtitles and for citing a specific moment. A transcription service gives you running text that reads like a document. Pick based on what the text is for. If you need subtitles, use the caption file. If you are writing an article, study notes or an interview write-up, use running text.
Common problems and how to fix them
"Show transcript" is missing
The video has no captions, or the creator turned them off. Automatic captions can also take a while to appear on newly uploaded videos, and they are not generated for every language or every recording. Check whether the CC button is shown on the player. If it is not, there is no transcript on YouTube to copy, and method 4 is the only option.
The transcript is in the wrong language
Use the language selector at the bottom of the transcript panel. If the language you want is not listed, the video has no captions in that language. Auto-translate shows translated captions on the player, but the result is a translation of the captions, not a transcript of the speech.
YouTube Shorts do not show a transcript
The Shorts player has fewer controls than the normal one. Open the same video in the regular player by changing /shorts/ in the address to /watch?v=, keeping the video ID, and look for the transcript there.
Live streams and premieres
Captions shown during a live stream are not always saved as a transcript right away. The full transcript usually appears once YouTube has processed the recording after the stream ends.
Music videos and videos with heavy background sound
Speech recognition struggles with singing and dense mixes, and that applies to YouTube's automatic captions too. If you are after lyrics, read how to get song lyrics from an audio file for what works and what does not.
The copied text is one long block
That is normal for automatic captions. Add paragraph breaks every time the topic changes and fix the names first. The practical guide to converting speech to text covers a three-pass cleanup that works for any machine transcript.
Using transcripts of other people's videos
Using a transcript for your own study, research, notes or accessibility is ordinary use. Quoting a short passage with attribution and a link back to the video is normal practice too. Republishing someone else's full transcript as your own article is not, whatever tool produced it. The transcript is a working copy, not a licence to reuse the content.
Turning your own video into an article
If you want to repurpose your own video, the transcript is only the first step:
- Transcribe the full recording, using the original file for the best audio.
- Cut hard. Spoken language takes many more words than written language to say the same thing.
- Restructure around headings. A video runs in time order, while an article works best organised by topic.
- Check quotes against the video before you publish, especially numbers and names.
For interviews and conversations with several speakers, transcribing an interview without losing a day covers the editing order that saves the most time.
Frequently asked questions
Can I get a transcript of a YouTube video for free?
Yes. If the video has captions, the transcript is built into YouTube: click ...more under the video, then Show transcript. You can read and copy it at no cost. If the video has no captions, the SpeakToWords free plan lets you test transcription on 15 minutes of audio a month without a card. It shows a preview of each transcript, and the full text needs a trial or a paid plan.
How do I copy a YouTube transcript without timestamps?
Open the transcript panel, click the three-dot menu at its top and choose Toggle timestamps. The time codes disappear, and you can select the text and paste it into any document.
Why is there no "Show transcript" button on a video?
The video has no captions. The creator may have turned them off, automatic captions may not have been generated yet, or the spoken language may not be supported for automatic captions. In that case, transcribe the audio file instead.
Can I download a YouTube transcript as a file?
For your own videos, yes. In YouTube Studio go to Subtitles, choose the video and download the captions as SRT, VTT or SBV. For other people's videos, YouTube only lets you copy the transcript from the panel.
How accurate are YouTube's automatic transcripts?
Accuracy is good on clear audio with one speaker and drops with music, crosstalk, accents and technical terms. Automatic captions also tend to lack punctuation and never label speakers, so plan on editing before you quote or publish.
Can SpeakToWords transcribe a YouTube link directly?
No. SpeakToWords transcribes files you upload, such as MP4, MOV or MP3, up to 500 MB. For your own videos, upload the original file or download it from YouTube Studio first.
Ready to try it on a video of your own? Create a free account, upload the file and compare the result with YouTube's automatic captions. The difference shows within the first minute of text.