YouTube Transcripts Scraper

Get the full transcript of YouTube videos: every caption line with its timestamp, the whole text joined for reading or feeding to an AI model, and basic video details. You choose the language you prefer.

What you get

One row per video that has captions:

Field	Meaning
videoId, url, title	The video
channelName, channelId	Who published it
lengthSeconds, viewCount	Duration and views
language, languageName	The caption track returned, for example es / Spanish
isAutoGenerated	true when the captions are YouTube's automatic ones, false when the creator provided them
availableTracks	Every caption language the video offers
segmentCount	Number of caption lines
segments	Caption lines: startSeconds, durationSeconds, text
text	All caption lines joined into one string

Example row (shortened)

{
  "videoId": "dQw4w9WgXcQ",
  "title": "Rick Astley - Never Gonna Give You Up (Official Video) (4K Remaster)",
  "channelName": "Rick Astley",
  "lengthSeconds": 213,
  "language": "en",
  "languageName": "English",
  "isAutoGenerated": false,
  "segmentCount": 61,
  "segments": [
    { "startSeconds": 18.64, "durationSeconds": 3.24, "text": "♪ We're no strangers to love ♪" },
    { "startSeconds": 22.64, "durationSeconds": 4.32, "text": "♪ You know the rules and so do I ♪" }
  ],
  "text": "[♪♪♪] ♪ We're no strangers to love ♪ ♪ You know the rules and so do I ♪ ..."
}

Input

Parameter	Default	Description
videos	(required)	Video URLs or ids: watch, youtu.be, Shorts and embed links all work
language	en	Preferred caption language, for example es, de, id
proxyConfiguration	Residential	Keep residential (see below)

{ "videos": ["https://www.youtube.com/watch?v=aircAruvnKk", "youtu.be/dQw4w9WgXcQ"], "language": "es" }

How the language is chosen

1. Creator captions in your language
2. Automatic captions in your language
3. English (creator, then automatic)
4. The first track the video has

So a request for es on a video with only Korean captions returns the Korean ones. Check language on each row, and availableTracks for what else exists.

Good to know

• Use the residential proxy. YouTube answers most requests from Apify's own servers with "Sign in to confirm you're not a bot". Every video gets a fresh IP, and a blocked one is retried up to 4 times. Without a proxy most videos fail.
• You pay only for transcripts. A video without captions, private, removed or blocked is not saved as a row. It is listed in the run's status message instead, for example 12 of 13 transcript(s) saved. Failed: .... A run where no video works fails.
• Captions come from the video. Music videos often have lyric lines with ♪ marks, and some talks start with the translator's credit.
• Long videos are long. A 4-hour course gives about 1,900 lines and 270,000 characters of text.

Use cases

• Build a searchable library or an AI knowledge base from lectures, tutorials and podcasts.
• Summarise videos or turn them into articles and notes.
• Study what channels say about a topic over time.
• Read a video's captions in another language when the creator provided them.

Scope

The actor reads publicly available captions of public videos, the same ones any viewer can turn on. It uses no login and no YouTube account. Respect the rights of video creators and YouTube's terms when you use the text.