YouTube Transcript: Captions to Text with Timestamps avatar

YouTube Transcript: Captions to Text with Timestamps

Pricing

$2.00 / 1,000 transcript rows

Go to Apify Store
YouTube Transcript: Captions to Text with Timestamps

YouTube Transcript: Captions to Text with Timestamps

Paste YouTube links or ids and get each video's transcript from its public captions: full text, timestamped segments, language (manual, automatic or translated), title, channel, length, available languages. Proxy option for rate limits. Videos without captions never charged.

Pricing

$2.00 / 1,000 transcript rows

Rating

0.0

(0)

Developer

Luigi Fiano

Luigi Fiano

Maintained by Community

Actor stats

0

Bookmarked

3

Total users

2

Monthly active users

10 days ago

Last modified

Categories

Features

Limited permissions

Share

Paste YouTube links or video ids and get one row per video with its transcript from the public captions: the full text, the segments with start time and duration, the caption language (written by the channel, automatic, or translated by YouTube), plus the title, channel, length and every caption language the video has.

What it is for

  • Summaries and notes with an LLM: feed the full text of a talk, a lesson, a webinar or a podcast episode.
  • Content repurposing: quotes, blog posts, show notes and newsletters from your own videos.
  • Search and RAG: index what is said in a list of videos, with timestamps that link back to the exact moment (url + &t= + start).
  • Language work: captions in the language you choose, or translated by YouTube.

How it works

  • For each video the public player data is read (title, channel, length, caption tracks), then the chosen caption track. Two or three requests per video: no browser, no video download.
  • Language: languages in order of preference; captions written by the channel come before YouTube's automatic ones (preferManualCaptions). If the video has none of your languages, its first caption language is used (fallbackToAnyLanguage) and usedFallbackLanguage is true.
  • Translation: with translateTo, YouTube's own machine translation of the captions is returned, unless the video already has captions in that language.
  • Nothing else is read: no comments, no likes, no data about viewers.
  • Private, age-restricted, members-only and unavailable videos are reported in the free ERRORS record, never bypassed.

When YouTube limits requests: the proxy option

YouTube limits how many requests one IP address can make, and it may limit the data-center addresses that Actors run on. When that happens, the video goes to the free ERRORS record with one of these reasons:

  • YouTube is limiting requests from this IP (HTTP 429)
  • YouTube asks to sign in to confirm this is not a bot
  • YouTube returned an empty transcript (it may be limiting requests from this IP)

Run those videos again with the Proxy option on (Apify Proxy). Every video then gets its own proxy IP, and a new one each time YouTube limits it, up to maxRetries times. Without the proxy, the Actor makes direct requests.

Input

FieldWhat it doesDefault
videosOne video id or link per line (watch, youtu.be, shorts, embed, live, m. and music. links)required
languagesCaption languages in order of preference["en"]
preferManualCaptionsCaptions written by the channel before automatic onestrue
allowAutoGeneratedUse automatic captions when there are no otherstrue
fallbackToAnyLanguageUse the video's first caption language when yours are missingtrue
translateToLanguage code for YouTube's machine translation (empty: none)empty
includeTimestampsAdd the segments with start time and durationtrue
proxyConfigurationApify Proxy, for when YouTube limits requestsoff
concurrency, timeoutSecs, maxRetriesSpeed and resilience3, 30, 2
maxVideos, maxResultsSafety caps (0: no limit on rows)1000, 0
{
"videos": [
"https://www.youtube.com/watch?v=dQw4w9WgXcQ",
"https://youtu.be/jNQXAC9IVRw",
"arj7oStGLkU"
],
"languages": ["en", "es"],
"includeTimestamps": true
}

Output

One row per video. This row comes from the Actor's own test, on a sample video:

{
"type": "transcript",
"videoId": "Ex4mpleVid0",
"url": "https://www.youtube.com/watch?v=Ex4mpleVid0",
"title": "How to brew pour-over coffee at home",
"channel": "Example Coffee Channel",
"channelId": "UCxxxxxxxxxxxxxxxxxxxxxx",
"durationSeconds": 412,
"duration": "6:52",
"language": "en",
"languageName": "English",
"captionsLanguage": "en",
"isAutoGenerated": false,
"isTranslated": false,
"usedFallbackLanguage": false,
"availableLanguages": [
{ "code": "en", "name": "English", "autoGenerated": false },
{ "code": "en", "name": "English (auto-generated)", "autoGenerated": true },
{ "code": "es", "name": "Spanish", "autoGenerated": false }
],
"text": "Today we brew a pour-over at home. You need fresh beans and a burr grinder. Heat the water to about 94 degrees.",
"wordCount": 23,
"characterCount": 111,
"segmentsCount": 3,
"segments": [
{ "start": 0.4, "duration": 3.1, "text": "Today we brew a pour-over at home." },
{ "start": 3.5, "duration": 2.8, "text": "You need fresh beans and a burr grinder." },
{ "start": 6.3, "duration": 3.4, "text": "Heat the water to about 94 degrees." }
],
"input": "https://www.youtube.com/watch?v=Ex4mpleVid0",
"error": null,
"scrapedAt": "2026-11-09T05:00:00.000Z"
}
  • language is the language of the text you get; with a translation it is the target language and captionsLanguage is the language of the original captions.
  • segments is null when includeTimestamps is off; the full text is always there.
  • channel and channelId are the public name and id of the channel that published the video.

Videos that cannot be read (not charged)

A video without captions, private, age-restricted, unavailable or limited by YouTube never becomes a dataset row, so it costs nothing. It is listed with the reason in the ERRORS record of the key-value store (also in the Output tab):

{ "type": "error", "videoId": "zzzzzzzzzz0", "input": "https://youtu.be/zzzzzzzzzz0", "url": "https://www.youtube.com/watch?v=zzzzzzzzzz0", "title": null, "error": "video unavailable (Video unavailable)", "scrapedAt": "2026-11-09T05:00:00.000Z" }

Other reasons you can see: the video has no captions (no subtitles and no automatic captions), no captions in the requested language(s): fr. Available: en, en (auto-generated), it, private video: nothing was bypassed, YouTube asks to sign in to watch this video (age-restricted or members-only): nothing was bypassed, not a YouTube video link or id, plus the three limits listed in the proxy section.

Summary

The SUMMARY record has the run totals: videosChecked, videosOk, videosWithError, limitedByYouTube, proxyRetries, rows, words and stoppedEarly.

Pricing

Pay per result: $0.002 per transcript row in the dataset ($2 per 1,000 videos). Videos without captions or that cannot be read are not charged: they go to the free ERRORS record. Use maxResults to cap the cost of a run.

Tips

  • Only need the text? Turn off includeTimestamps: rows get much smaller.
  • Big lists: raise concurrency (up to 10) and the run timeout in the run options. If many videos come back limited, run them again with the proxy on.
  • Shorts and live replays work when they have captions.

Notes

  • Only public data: the public player data and the public captions. No login, no cookies, no comments, no data about viewers.
  • You are responsible for how you use the transcripts (copyright and YouTube's terms).
  • Not affiliated with YouTube or Google. YouTube is a trademark of Google LLC.