YouTube Transcript Scraper

Extract transcripts from YouTube videos, playlists, and entire channels — as clean full text, timestamped segments, and ready-to-use SRT subtitles.

The difference: this actor also handles videos that have no captions. Most transcript scrapers return an error the moment a video has no subtitle track. This one transcribes the audio with speech recognition instead, so you get a usable transcript either way.

Quick start

Paste a video URL and run it. That is the whole setup.

{
  "videoUrls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"]
}

Or point it at a whole channel or playlist:

{
  "channelUrl": "https://www.youtube.com/@mkbhd",
  "maxVideos": 25,
  "languages": ["en"],
  "includeSrt": true
}

What you get back

Field	Description
transcript	The full transcript as clean, continuous text
segments	Timestamped segments — start, end, duration, text
srt	Ready-to-use SRT subtitle content (optional)
language	Language code of the transcript
transcriptSource	youtube-captions, youtube-auto-captions, or speech-recognition
wordCount	Number of words in the transcript
title · channel · channelId	Video and channel metadata
durationSeconds · viewCount	Video statistics
availableLanguages	Every caption language the video offers
scrapedAt	ISO timestamp of extraction

The transcriptSource field always tells you where the text came from, so you know whether you are reading human-written captions, YouTube's auto-captions, or speech recognition.

Example output

{
  "videoId": "dQw4w9WgXcQ",
  "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "title": "Rick Astley - Never Gonna Give You Up (Official Video)",
  "channel": "Rick Astley",
  "durationSeconds": 213,
  "language": "en",
  "transcriptSource": "youtube-captions",
  "wordCount": 486,
  "transcript": "♪ We're no strangers to love ♪ ♪ You know the rules and so do I ♪ ...",
  "segments": [
    { "start": 18.64, "duration": 3.28, "end": 21.92, "text": "♪ We're no strangers to love ♪" }
  ],
  "availableLanguages": ["en", "de-DE", "ja", "pt-BR", "es-419"],
  "scrapedAt": "2026-07-31T09:47:39.257Z"
}

Input options

Option	Default	What it does
videoUrls	[]	Video URLs or IDs. Accepts watch links, youtu.be, Shorts, embeds, or bare 11-character IDs
playlistUrl	—	Transcribes every video in a playlist
channelUrl	—	Transcribes a channel's most recent uploads. Works with @handle, /channel/, /c/, and /user/ URLs
maxVideos	50	Hard cap on videos per run
languages	["en"]	Language codes in order of preference. First available match wins
preferManualCaptions	true	Prefer human-written captions over auto-generated ones — they are noticeably more accurate
asrFallback	true	Transcribe the audio when no captions exist at all
includeSegments	true	Return timestamped segments alongside the full text
includeSrt	false	Also return an SRT subtitle string

You can combine videoUrls, playlistUrl, and channelUrl in a single run. Duplicates are removed automatically.

What people use it for

• Content repurposing — turn long videos into blog posts, newsletters, and social clips
• AI and RAG pipelines — feed video content into embeddings, summarisation, or Q&A
• Research at scale — search across hundreds of videos by what was actually said
• Subtitling and localisation — export SRT, then translate
• SEO — publish transcripts so video content becomes indexable text
• Accessibility — provide text alternatives for video content
• Competitive and market research — track how topics are discussed across channels

Pricing

Pay per transcript. You are charged only for transcripts actually delivered — videos that are private, deleted, region-blocked, or otherwise fail cost you nothing. Failed videos still appear in the dataset with an error field explaining why, so nothing disappears silently.

Reliability notes

• Videos are processed independently. One failure never aborts the run.
• Requests are retried automatically on transient network and rate-limit errors.
• Private, deleted, and age-restricted videos cannot be transcribed.
• Live streams can only be transcribed once they have finished processing.
• Speech recognition is used only when a video has no captions in any requested language. It is accurate for clear speech and less so over heavy background music or overlapping speakers.
• Very long videos are truncated when falling back to speech recognition.

Legal

This actor reads publicly available data only. It does not access private videos, bypass age restrictions, or require you to provide YouTube credentials. You are responsible for ensuring your use of the extracted data complies with YouTube's Terms of Service and with copyright law in your jurisdiction.

Support

Missing a field, need another output format, or hit a video that should work but does not? Open an issue on the Issues tab of this actor. Reports that come with the video URL are usually fixed quickly.