How to Transcribe a YouTube Video: 3 Ways That Work

By whispertranscribe.ai Editorial Team · Last updated: November 2026 · 12 min read

Here's how to transcribe a YouTube video in three ways: open YouTube's built-in transcript if the video has captions, copy those captions into a document and clean them up, or paste its link into an AI transcription tool. Only the third method still works when a video has no captions at all.

This guide covers all three, step by step. You will also find a cleanup checklist for copied YouTube transcripts, a table to help you pick a method, and a workflow for turning one transcript into subtitles, a summary, or a blog post.

Quick answer: how to transcribe a YouTube video

The fastest free route is YouTube's own Show transcript panel, but it only appears when a video has captions and the text can't be downloaded as a file. If you need a clean, exportable YouTube transcript, or the video has no captions, use an AI tool that accepts a link.

  1. Check for a transcript on YouTube. Open the description and click Show transcript.
  2. Copy and clean it. Paste the text into a doc and strip out timestamps and sound tags.
  3. No transcript, or need a file? Paste the link into whispertranscribe.ai's URL transcription tool and export TXT, DOCX, or SRT.

Method 1: Use YouTube's built-in "Show transcript"

YouTube shows a full transcript for any video that has captions, either captions the creator uploaded or captions YouTube generated automatically. According to YouTube Help, you open it from the video description, and the transcript scrolls along as the video plays.

What you need: a video with captions. If the creator hasn't added captions and YouTube hasn't generated them, the button won't appear.

On desktop (web browser)

  1. Open the video on youtube.com.
  2. Click the description below the video to expand it.
  3. Scroll down to the end of the description and click Show transcript. A transcript panel opens next to the player.
  4. Click any line to jump to that moment in the video.
  5. To copy the text, select it in the panel and paste it into a document. The panel's menu includes a Toggle timestamps option, so you can hide timestamps before you copy.

On mobile (YouTube app)

  1. Open the video in the YouTube app.
  2. Tap the video description to open it.
  3. Tap Show transcript, then tap any line to jump to that point.

YouTube's help page covers reading the transcript and jumping around in it. It doesn't mention copying text out of the app. If you need the text in a document, the desktop site is easier.

💡 Tip: If the video is yours, you don't need the transcript panel at all. In YouTube Studio, go to Subtitles, open the video, click Edit next to the language, then Options → Download subtitles (YouTube Help). Viewers don't get this option.

Limitations of YouTube's own transcript

YouTube's transcript is good for a quick read or finding a quote. It falls short when you need complete, accurate text you can reuse. Every limitation below is based on YouTube's own help documentation.

1. No captions means no transcript

The transcript panel is built from the caption track. If a video has no captions, there is nothing to show. YouTube lists several reasons automatic captions might not be generated: the audio is still processing, the language isn't supported, the video is too long, the sound quality is poor, there is a long silence at the start, or speakers overlap or switch languages (Use automatic captioning).

2. Automatic captions can get words wrong

YouTube says plainly that automatic captions "might misrepresent the spoken content due to mispronunciations, accents, dialects, or background noise," and recommends that creators review them. Unless the creator fixed the captions, whatever the speech recognition got wrong ends up in your transcript.

3. Automatic captions only cover one language

If a video has multiple audio tracks, YouTube generates automatic captions in the default language only. Videos where speakers mix two languages are also on YouTube's list of cases where automatic captions may fail.

4. No speaker labels

Automatic captions turn speech into text. YouTube's help pages don't describe any speaker identification, so in an interview or a panel discussion, you get one long stream of lines with no indication of who said what. Names show up only if the creator typed them into the captions.

5. Viewers can't download a file

As a viewer, you can read, click, and copy. There's no download button for a TXT, DOCX, or SRT file. The Download subtitles option is in YouTube Studio, so only the channel owner (or someone with Studio access) can use it.

Method 2: Transcribe a YouTube video with whispertranscribe.ai

whispertranscribe.ai is a browser-based transcription tool built on OpenAI's Whisper speech recognition model. You paste a public YouTube, Vimeo, Google Drive, or Dropbox link, and it transcribes the audio itself. It doesn't depend on existing captions, so it works on videos YouTube never captioned.

Whisper was trained on 680,000 hours of multilingual and multitask supervised data, according to OpenAI's research paper, which reports that the models generalize well without fine-tuning. whispertranscribe.ai supports 130+ languages and reaches up to 99% accuracy on clear audio. Background noise, heavy accents, and crosstalk still reduce accuracy, so plan to review the result.

Step 1: Open the URL tool and pick the spoken language

Go to Transcribe from URL (or the YouTube transcription page). In Step 1, choose the language spoken in the video. Choose a regional variant if there is one, such as English (UK), English (India), Spanish (Latin America), or Portuguese (Brazil).

Copy the video URL from your browser's address bar or YouTube's Share button, and paste it into the link field. The link has to be publicly accessible.

Step 3: Turn on Speaker Detection (optional)

If more than one person talks, as in an interview, podcast, or panel, switch on Enable Speaker Detection. The transcript will then label each speaker, which is the piece YouTube's own transcript lacks.

Step 4: Start the transcription

Click Start URL Transcription. You'll get a text transcript with timestamps.

Step 5: Export in the format you need

Export as TXT or DOCX for notes and editing, PDF to share, CSV to analyze, or SRT/VTT for subtitle files.

You can start transcribing for free with no account required. Paid plans start from $9.99/month billed annually; see pricing for current details.

Method 3: Upload the video file (when it's your own video)

If you made the video, you probably have the original file, and uploading that file is usually the cleanest option. You skip any problems with the link, and the audio is the best version you have.

  1. Go to the whispertranscribe.ai homepage (Transcribe Files).
  2. Choose the spoken language.
  3. Drag and drop your file, or browse to it. Supported formats include MP4, MOV, MKV, WEBM, AVI, MP3, WAV, M4A, and more, and single files can be up to 5GB.
  4. Turn on Speaker Detection if needed, then start transcribing and export.

Only upload files you have the right to use: videos you created, footage you licensed, or content you have permission to transcribe. This guide won't cover downloading other people's YouTube videos. YouTube's Terms of Service don't allow downloading content unless YouTube expressly allows it or you have permission. For someone else's public video, use Method 1 or Method 2.

Is the file mostly talking and very large? You can convert it to MP3 first to make the upload smaller.

Which method should you choose?

Use YouTube's transcript when you just need to read or quote a few lines. Use a link-based AI tool when the video has no captions, has several speakers, or you need a file. Upload the original file when the video is yours.

What you need Best method Why
Read along or find a quote quickly Method 1: YouTube Show transcript Free and built in, no extra tools
Text from a video with no captions Method 2: Paste the link Transcribes the audio, not the captions
Labels for who said what Method 2 or 3 with Speaker Detection YouTube captions carry no speaker labels
A downloadable TXT, DOCX, or PDF Method 2 or 3 YouTube offers viewers no download button
SRT or VTT subtitles Method 2 or 3, export SRT/VTT Subtitle files with timing included
Transcript of your own video Method 3: Upload the original file Best audio quality, no link needed
A video that mixes languages Method 2 with Multilingual Mix YouTube auto captions use one language
A lecture or stream several hours long Method 2 or 3 Plans support files up to 10 hours

How to clean up a copied YouTube transcript

If you go with Method 1, the copied text needs some work before you can use it. Run through this checklist:

  1. Remove timestamps. Click Toggle timestamps in the transcript panel before copying. If they've already been pasted, delete any line that contains only a timestamp. In a text editor that supports regular expressions, find ^\d{1,2}(:\d{2}){1,2}\s*$ and replace it with nothing.
  2. Join the broken lines. Captions are split into short screen-sized chunks. Replace single line breaks with spaces, then add paragraph breaks wherever the topic changes.
  3. Delete sound tags. English automatic captions can include sounds and noises in brackets (YouTube calls these expressive captions). Search for [ and remove tags like [Music] or [Applause] unless you need them.
  4. Fix the capitalization. Expressive captions may show loud speech in all caps. Change it back to normal case.
  5. Add punctuation and speaker names. Auto captions often have little punctuation, and they never say who's speaking. Listen again for the sections you plan to quote.
  6. Check names and numbers. Before you publish anything, compare proper nouns, figures, and technical terms against the video.

If the list looks long, Method 2 handles most of it for you: you get speaker labels and timestamps you can keep or drop on export.

Turn one transcript into subtitles, a summary, or a blog post

A transcript is useful after transcription, too. Here is a simple workflow:

  • Subtitles: export SRT or VTT and upload the file to YouTube Studio under Subtitles → Upload file. For more control, use the AI subtitle generator, or the MP4 to SRT converter if you have the video file.
  • Summaries and notes: export TXT and open it in the Whisper Transcribe AI Summary GPT in ChatGPT to get key points, chapter notes, or a study guide.
  • Blog posts and show notes: start from the DOCX export, keep the strongest quotes, and rewrite the rest as an article. Transcripts give you the raw material, not a finished post.
  • Translation: if the video isn't in English and you need English text, use translate audio to English.

Long videos and multilingual videos

Two cases where YouTube's transcript often doesn't help:

  • Long videos. Conference recordings, livestream replays, and full-day courses are where auto captions tend to give out, and YouTube lists "the video is too long" as one reason captions don't appear. whispertranscribe.ai plans support files up to 10 hours, with single uploads up to 5GB.
  • Multilingual videos. Interviews that switch between, say, English and Spanish confuse single-language captions. whispertranscribe.ai's Multilingual Mix mode handles recordings that mix English, Spanish, French, German, Hindi, Russian, Portuguese, Japanese, Italian, and Dutch.

Tips for better results

  • Pick the exact language variant. English has US, UK, Australia, New Zealand, and India options, and Chinese has Simplified, Traditional, and Cantonese.
  • Use Noise / Low-Quality mode for rough audio. Many languages have a dedicated option for noisy or low-quality audio and video, such as street interviews, phone footage, and old uploads.
  • Use Multilingual Mix for code-switching. Choose it when speakers switch languages mid-conversation.
  • Try English (Songs & Lyrics) for music. It's a separate option for sung vocals.
  • Turn on Speaker Detection for multiple voices. It makes interviews and panels far easier to read.
  • Try a different option if the result is off. whispertranscribe.ai's own guidance notes that some languages have multiple options. If one result isn't good enough, rerun with another.
  • Upload the original when you can. The original file of your own video gives the model the best audio to work with.

For other kinds of footage, see the general video to text converter.

Only transcribe content you have the right to use. A transcript of someone else's video is still their words. Quoting short passages with credit for commentary, research, or teaching is often fine, while republishing a full transcript usually requires the creator's permission. YouTube explains how fair use works in its fair use guide, and the rules differ from country to country. This section is general information, not legal advice. If you're unsure, ask the creator or a lawyer.

FAQs

Can I transcribe a YouTube video without captions?

Yes. Use a tool that transcribes the audio rather than reading the caption track. YouTube's Show transcript panel only appears when a video has captions, so it can't help with uncaptioned videos. With whispertranscribe.ai, you paste the public YouTube link, choose the spoken language, and the speech is converted to text from the audio. Because it doesn't rely on captions, it also works when YouTube's automatic captions failed because of noise, length, or mixed languages.

Can I transcribe a YouTube video for free?

Yes. YouTube's built-in transcript is free for any video with captions. Open the description, click Show transcript, and copy the text. It doesn't give you a downloadable file or speaker labels, and it won't work without captions. whispertranscribe.ai also has a free starting point that requires no account, so you can test a video before deciding whether you need a paid plan. Paid plans start from $9.99/month billed annually.

How do I get the transcript of a YouTube video without timestamps?

Hide the timestamps before you copy. On desktop, open Show transcript, open the transcript panel's menu, and choose Toggle timestamps. Then select and copy the text. If timestamps have already been pasted into your document, delete the timestamp-only lines with a find-and-replace (see the cleanup checklist above). With whispertranscribe.ai, export TXT or DOCX for readable text, or SRT/VTT if you do want timing data.

Can I transcribe a YouTube video in another language or translate it to English?

Yes. whispertranscribe.ai supports 130+ languages, including regional variants such as Spanish (Latin America) and Cantonese. Choose the spoken language before you start. If speakers switch languages, choose Multilingual Mix. If you need English text from a non-English video, use the translate audio to English tool. YouTube's automatic captions cover a shorter list of languages and are generated only in a video's default language.

How long can a YouTube video be for transcription?

whispertranscribe.ai plans support files up to 10 hours long, and single uploads can be up to 5GB. That covers most lectures, podcasts, and livestream replays in one go. On YouTube's side, very long videos sometimes don't get automatic captions at all, which is one reason the transcript panel may be missing on long recordings. For multi-hour content, a link- or file-based tool is the more reliable option.

Can I save a YouTube transcript as a TXT or Word file?

Not from YouTube itself, if you are a viewer. YouTube's transcript panel lets you read and copy text, but it has no download button. If you own the video, YouTube Studio lets you download the caption track under Subtitles → Options → Download subtitles. For any public video, paste the link into whispertranscribe.ai and export the transcript as TXT, DOCX, PDF, CSV, SRT, or VTT.


Try it on your next video: paste a YouTube link into whispertranscribe.ai to get a transcript in 130+ languages, with speaker labels and SRT/VTT export.

References

Ready to turn your recordings into text?

Transcribe audio and video with whispertranscribe.ai.

Start transcribing ↗