YouTube Transcript API & Scraper

What is YouTube Transcript API & Scraper?

YouTube Transcript API & Scraper is a YouTube transcript API and transcript extractor: it gets the transcript of any YouTube video that has captions, captions and subtitles alike: the full text in one field, plus timestamped segments (start, duration, text) for every caption line. Paste watch URLs, youtu.be links, Shorts URLs, or bare video ids, in any mix, or give it channels and playlists, and choose your preferred caption languages. Human-made captions are used first; YouTube's auto-generated captions are the fallback, and you can switch them off.

Each video also comes back with its title, channel, channel id, duration, publish date, view count, the caption language used, whether it was auto-generated, and the list of every caption language the video offers.

You pay per transcript delivered: $0.005 per transcript ($5 per 1,000), platform usage included. Videos without captions, unavailable videos, and blocked requests are reported in the dataset and free.

The easiest way to try it: paste a few video links into the Videos field and press Start.

What data can you extract from YouTube videos?

Field	Example
videoId, url	iG9CE55wbtY, https://www.youtube.com/watch?v=iG9CE55wbtY
title	Do schools kill creativity? | Sir Ken Robinson | TED
channelName, channelId	TED, UCAuUUnT6oDeKwE6v1NGQxug
durationSeconds	1203
publishDate	2007-01-06T21:00:34-08:00
viewCount	25037463
language, languageName	en, English
isAutoGenerated	false (human-made captions)
isAutoDubbed	false (see Limits)
matchedPreferredLanguage	true
availableLanguages	every caption track: [{ languageCode, name, isAutoGenerated, isAutoDubbed }]
transcriptText	the whole transcript as one string
segments	[{ "start": 1.36, "duration": 1.68, "text": "..." }] (seconds)
segmentCount, wordCount, characterCount	427, 3170, 17574
sourceUrl, sourceName	the channel or playlist a video came from (null for videos you listed yourself)
fetchStatus, error	ok, null

Example record (segments shortened):

{
  "videoId": "dQw4w9WgXcQ",
  "url": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "title": "Rick Astley - Never Gonna Give You Up (Official Video) (4K Remaster)",
  "channelName": "Rick Astley",
  "channelId": "UCuAXFkgsw1L7xaCfnd5JJOw",
  "durationSeconds": 213,
  "publishDate": "2009-10-24T23:57:33-07:00",
  "viewCount": 1826397375,
  "language": "en",
  "languageName": "English",
  "isAutoGenerated": false,
  "isAutoDubbed": false,
  "matchedPreferredLanguage": true,
  "availableLanguages": [
    { "languageCode": "en", "name": "English", "isAutoGenerated": false, "isAutoDubbed": false },
    { "languageCode": "en", "name": "English (auto-generated)", "isAutoGenerated": true, "isAutoDubbed": false }
  ],
  "transcriptText": "[♪♪♪] ♪ We're no strangers to love ♪ ♪ You know the rules and so do I ♪ ...",
  "segments": [
    { "start": 1.36, "duration": 1.68, "text": "[♪♪♪]" },
    { "start": 18.64, "duration": 3.24, "text": "♪ We're no strangers to love ♪" }
  ],
  "segmentCount": 61,
  "wordCount": 487,
  "characterCount": 2089,
  "input": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "fetchedAt": "2026-10-14T04:50:12.306Z",
  "fetchStatus": "ok",
  "error": null
}

Why use this YouTube transcript scraper?

• Every video comes back. Each input line produces a record, including failures, with a fetchStatus and an error message. Nothing is dropped silently, and failed records are free.
• Built for YouTube's bot check. YouTube answers many cloud IPs with "Sign in to confirm you're not a bot". When that happens the actor moves to a fresh Apify Proxy session and keeps the sessions YouTube lets through, instead of failing the video.
• Tested before release. In the release verification runs on Apify (2026-10-13), 17 of 17 captioned test videos returned full transcripts: human and auto-generated captions; English, Spanish, German, Japanese, Korean and French; 19 seconds to 3 hours 37 minutes; a Short, music videos, and uploads from the same day. A 60-video run over ten channels returned transcripts for all 58 videos that had captions. The video without captions, the removed video, and the malformed input came back as free failure records. A channel run (Veritasium, 10 newest videos) and a playlist run (10 videos) listed and transcribed 20 of 20.
• Original speech first. On videos with YouTube's AI auto-dubbing, auto-generated tracks in dubbed languages transcribe the dub, not what was said. The actor prefers the original-language track and flags dub tracks with isAutoDubbed.
• No browser, no sign-in. Plain HTTP requests, no YouTube account and no cookies.

How to scrape YouTube transcripts

1. Open the Input tab.
2. Add videos to Videos, one per line: https://www.youtube.com/watch?v=..., https://youtu.be/..., https://www.youtube.com/shorts/..., or an 11-character id. Or add Channels or Playlists and set Max videos per channel or playlist.
3. Set Preferred languages (default en). Add more codes in order of preference, for example es, de.
4. Leave Allow auto-generated captions on unless you need human-made captions only.
5. Leave Proxy configuration on Apify Proxy.
6. Press Start. When the run finishes, open the Output tab or download the dataset as JSON, CSV, or Excel.

Call the YouTube transcript API from code

Call the actor over the Apify API and get the transcripts back in the same request:

curl -X POST "https://api.apify.com/v2/acts/morgan-studio~youtube-transcript-scraper/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"videoUrls": ["https://www.youtube.com/watch?v=jNQXAC9IVRw"], "preferredLanguages": ["en"]}'

From Python, with the apify-client package:

from apify_client import ApifyClient

client = ApifyClient("YOUR_APIFY_TOKEN")
run = client.actor("morgan-studio/youtube-transcript-scraper").call(
    run_input={"videoUrls": ["https://www.youtube.com/watch?v=jNQXAC9IVRw"]}
)
for item in client.dataset(run["defaultDatasetId"]).iterate_items():
    print(item["title"], item["transcriptText"][:200])

The requests run on Apify's servers, not your IP, so a blocked IP on your side doesn't stop them. The API tab above has snippets for JavaScript, the CLI and MCP.

YouTube transcript extractor for channels and playlists

Put channel URLs in Channels or playlist URLs in Playlists and set Max videos per channel or playlist. Each video comes back as its own record, so a whole playlist's transcripts export in one run.

Download YouTube transcripts as JSON, text or CSV

Each record carries the full transcript in transcriptText and the timestamped lines in segments. Download the dataset from the Output tab as JSON, CSV or Excel, or from https://api.apify.com/v2/datasets/<datasetId>/items?format=json.

Input

• Videos (videoUrls): watch, youtu.be, Shorts, embed or live URLs, or bare 11-character ids. Duplicates are skipped.
• Channels (channelUrls): https://www.youtube.com/@handle, /channel/UC..., /c/... or /user/.... The newest videos on the Videos tab; add /shorts or /streams for those tabs.
• Playlists (playlistUrls): any URL with ?list=..., in playlist order.
• Max videos per channel or playlist (maxVideosPerSource): default 20, up to 1,000.
• Preferred languages (preferredLanguages): caption language codes in order, default ["en"]. For each language a human-made track is used first, then an auto-generated one. If none exist, the actor returns the video's own track (human-made first) and sets matchedPreferredLanguage to false.
• Allow auto-generated captions (allowAutoGenerated): default on.
• Include timestamped segments (includeTimestamps): default on. Off returns transcriptText only.
• Max concurrency (maxConcurrency): videos fetched at once, default 3, up to 10.
• Proxy configuration (proxyConfiguration): default Apify Proxy. Strongly recommended.

Example input:

{
  "videoUrls": [
    "https://www.youtube.com/watch?v=iG9CE55wbtY",
    "https://youtu.be/jNQXAC9IVRw",
    "https://www.youtube.com/shorts/aNI2jv7lBPQ",
    "kJQP7kiw5Fk"
  ],
  "preferredLanguages": ["en", "es"],
  "allowAutoGenerated": true,
  "includeTimestamps": true
}

How much does it cost to scrape YouTube transcripts?

You pay per transcript delivered. There is no separate charge for Apify platform usage; it is included in the price.

Event	Price
Transcript (fetchStatus: "ok")	$0.005 ($5 per 1,000)
Actor start	$0.00005 per run

Only videos that came back with a transcript are charged. Videos without captions, unavailable or private videos, blocked requests, and invalid input are free. They are still written to the dataset so you can see what happened.

For example, 200 videos where 180 have captions cost 180 × $0.005 = $0.90.

You can cap a run's total charge in the run options (maximum cost per run). When the next transcript would pass the cap, the run stops and logs how many videos were not processed.

Limits

• Videos without captions have no transcript. The actor reads YouTube's caption tracks; it does not run speech-to-text. Such videos come back as no-captions (free).
• YouTube changes this path; failures are free. The actor uses the caption data YouTube's mobile apps receive. YouTube changes these internals without notice, and tightens its bot check from time to time. When a change breaks a request, the affected videos come back as error or blocked records and are not charged.
• Use the proxy. Without Apify Proxy, most cloud runs get the bot check and the videos come back as blocked.
• Auto-dubbed tracks. YouTube's AI dubbing adds auto-generated caption tracks in the dubbed languages. These are listed in availableLanguages with isAutoDubbed: true and used only when no original-language track fits.
• Age-restricted, members-only, and private videos need a signed-in account; they come back as login-required or private (free).
• Channel and playlist listing reads the same pages a signed-out visitor sees. A channel or playlist that cannot be listed comes back as one source-not-found or source-error record (free).

How failures are reported

Every input line produces one record. Each record has a fetchStatus:

• ok: the transcript was read.
• no-captions: the video has no caption tracks (or only auto-generated ones while those are switched off).
• unavailable, private, login-required: YouTube will not play this video to a signed-out visitor.
• blocked: YouTube kept answering with its bot check on 10 proxy sessions in a row.
• invalid-input: the line is not a YouTube video, channel, or playlist URL or id.
• source-not-found, source-error, source-empty: a channel or playlist could not be listed, or lists no videos.
• error: the caption download failed repeatedly, or the response could not be read.

The error field says what happened. The run log ends with a summary line: transcripts, failures by kind, bot checks seen, proxy sessions retired, requests, elapsed time, and usage.

FAQ

What about "YouTube transcript API IP blocked" errors? Those come from fetching transcripts from a cloud server's IP. This actor runs through Apify Proxy and retries on a fresh session when YouTube answers with a bot check, so your own IP never touches YouTube.

How do I export a YouTube playlist's transcripts? Paste the playlist URL into Playlists, set how many videos to take, run, and download the dataset. One record per video.

Can I get the transcript as JSON? Yes. Every record is JSON, with transcriptText (one string) and segments (start, duration, text). Switch off Include timestamped segments if you only want the text.

Does it need a YouTube API key or a login? No. It reads the captions YouTube shows to a signed-out viewer.

Which languages are supported? Any language a video has captions in. Put the codes you want in Preferred languages, in order. availableLanguages lists every track the video offers.

Can it transcribe videos that have no captions? No. This actor returns YouTube's own captions only. Videos without them are reported as no-captions and not charged.

Can I call it from my code or an AI agent? Yes. Like any Apify actor it runs through the Apify API: start a run with your input and read the dataset as JSON. transcriptText is ready to pass to a language model.

Can I transcribe a whole channel or playlist? Yes. Put it in Channels or Playlists and raise Max videos per channel or playlist. Each video is charged only if it has a transcript.

How do I keep costs under control? Set a maximum cost per run in the run options. You are only charged for transcripts actually delivered.

Is it legal to scrape YouTube transcripts? The actor reads only data that a signed-out visitor can see. Whether your use is allowed depends on what you do with the data and where you are. If you are unsure, ask a lawyer.

Ready-made examples

Each one opens with its input filled in; copy it to your account and run.

• Download YouTube Shorts transcripts
• Export YouTube playlist transcripts
• YouTube channel transcripts: latest videos
• YouTube transcript text for AI and LLMs
• YouTube video transcripts with timestamps

More actors from the same developer

• Amazon Product Scraper & Price API (ASIN, Keyword)
• App Store Reviews Scraper: Apple & Google Play API
• Zillow Scraper API: Property, Search & Zestimate

Changelog

• 2026-10-15: README and listing reworded for search; new sections on API use, scheduling and export; links to the ready-made examples. No change to the scraper or the prices.
• 0.1 (2026-10-13): First release. Transcripts from video URLs and ids, channels, and playlists, preferred languages, human or auto-generated captions, timestamped segments, auto-dub detection, proxy session rotation on YouTube's bot check. Pay per transcript: $0.005, failures free.