YouTube Transcript & Subtitles Scraper, Original + Translated avatar

YouTube Transcript & Subtitles Scraper, Original + Translated

Pricing

from $2.80 / 1,000 transcripts

Go to Apify Store
YouTube Transcript & Subtitles Scraper, Original + Translated

YouTube Transcript & Subtitles Scraper, Original + Translated

Timestamped YouTube transcripts and subtitles from videos, channels or searches, in the real original language (auto-dubbed and non-English videos included) or translated. Videos without captions are flagged and free. Optional comments, new-video watch per channel, webhook.

Pricing

from $2.80 / 1,000 transcripts

Rating

0.0

(0)

Developer

Alexis Morain

Alexis Morain

Maintained by Community

Actor stats

0

Bookmarked

1

Total users

1

Monthly active users

4 days ago

Last modified

Features

Limited permissions

Share

What does YouTube Transcript & Subtitles Scraper do?

It returns the full timestamped transcript of YouTube videos: every caption line with its start and end time, plus the whole text in one field. You give it video URLs, channels or search queries. You get the transcript in the language actually spoken in the video, or translated into the language you choose.

It reads the caption track list of each video before asking for the text. That check is what makes the difference on three kinds of videos where a plain transcript request goes wrong:

  • Non-English videos. A transcript request that does not name the right language often comes back empty on a French, Spanish or Japanese video, exactly as if the video had no captions at all. This actor sees which tracks exist and asks for the right one.
  • Auto-dubbed videos. More and more videos carry an AI dub in English. A default request then returns the captions of the dub, not what the creator said. This actor finds the track marked as original and returns that one, unless you ask otherwise.
  • Regional tracks. A video whose only track is en-US comes back empty when you ask for en. The actor sends the exact track code.

Videos without any captions are flagged no_transcript and not charged.

Why use YouTube Transcript & Subtitles Scraper?

  • Feed an LLM or a RAG pipeline with what videos actually say: courses, podcasts, conference talks, product reviews.
  • Summarize or index a whole channel in one run, then keep it up to date with watch mode.
  • Research a topic across the top videos of a search, in several languages, translated to one.
  • Repurpose content: blog posts, show notes, quotes, subtitles for your own edits.
  • Monitor competitors or creators: new videos only, sent to your webhook.

You pay only for transcripts you receive. Videos without captions, unavailable videos, invalid lines and errors cost nothing.

What data can YouTube Transcript & Subtitles Scraper extract?

Transcript rows (type: transcript)

FieldExampleMeaning
videoId6p1S8B8wA70YouTube video id
urlhttps://www.youtube.com/watch?v=6p1S8B8wA70Watch URL
titleTrump prend une raclée...Video title
channelNameLe MondeChannel name
channelIdUC...Channel id
durationSeconds777Video length
viewCount160054895Views when scraped
statusokok, no_transcript, unavailable, language_unknown, timeout, invalid_input or error
languagefrCode of the caption track that was read
originalLanguagefrLanguage spoken in the video, or null when the video has no speech track to tell (music, text-only videos)
isOriginalLanguagetrueThe track read is in the original language (null when the original is unknown)
isAutoGeneratedtrueThe track is YouTube speech recognition, not captions uploaded by the creator
availableLanguages[{"code": "fr", "autoGenerated": true}]Every caption track the video has
translatedToenTarget of the automatic translation, or null
segments[{"start": 0, "end": 3.04, "text": "Les élections..."}]Timestamped lines, in seconds
textLes élections de mi-mandat...The whole transcript as one string
segmentCount404Number of lines
noteRequested language 'de' is not available; the original-language track was returned.Why a row is not what you asked for, when that happens
sourcechannelinput, channel or search
sourceValue@stanfordThe channel or query that led to the video
scrapedAt2026-11-09T15:00:00+00:00Time of the run

Comment rows (type: comment, option)

FieldMeaning
videoIdVideo the comment belongs to
authorComment author name
textComment text
publishedAtPublication time, ISO 8601 UTC
likesLike count
repliesReply count

How to scrape YouTube transcripts

  1. Paste video URLs, channel URLs or @handles, or search queries.
  2. Leave Transcript language on original, or type a language code such as en, fr or pt-BR.
  3. Optional: set Translate to to get every transcript in one language.
  4. Optional: turn on Include comments.
  5. Click Start. Download the dataset as JSON, CSV or Excel, or read it through the API.

Transcripts are fetched in batches. Expect one to three minutes before the first rows appear, whatever the number of videos.

How much does it cost to scrape YouTube transcripts?

EventFree and BronzeSilverGold and above
Transcript$3.50 per 1,000$3.00 per 1,000$2.80 per 1,000
Comment$0.50 per 1,000$0.50 per 1,000$0.50 per 1,000
Actor start$0.002 per run$0.002 per run$0.002 per run

Examples at the Free price: 100 transcripts cost $0.35. A channel of 500 videos costs $1.75. 100 videos with 100 comments each cost $0.35 + $5.00.

A translated transcript costs the same as an original one. Videos without captions, unavailable videos and errors are free. You can cap the spend of a run with Max transcripts or with the maximum charge setting of the run.

Free Apify plan: up to 0.15 USD of DataForSEO data per day; paid plans unlimited.

Input

FieldDefaultMeaning
videos[]Video URLs (watch, youtu.be, shorts, embed, live) or 11-character ids
channels[]Channel URLs, @handles or UC... ids; reads the Videos tab, most recent first
searches[]YouTube search queries; reads the video results
maxVideosPerSource20Videos read from each channel and each search, 1 to 500
languageoriginaloriginal, or a language code
translateToemptyLanguage code for automatic translation
includeCommentsfalseAdds comment rows for each video whose transcript was returned
maxComments100Comments per video, 20 to 700
monitorfalseWatch mode: per channel, only videos never delivered before
maxItems1000Most transcripts requested in one run
webhookUrlemptyReceives the run summary by POST

Example:

{
"videos": ["https://www.youtube.com/watch?v=UF8uR6Z6KLc"],
"channels": ["@stanford"],
"searches": ["intro to machine learning"],
"maxVideosPerSource": 10,
"language": "original",
"translateTo": "en"
}

Output

One dataset row per video, plus one row per comment when comments are on. The Transcripts view of the dataset shows the main fields; the Comments view lists comments. A short run summary (counts per status, delays) is saved in the key-value store under OUTPUT.

{
"type": "transcript",
"videoId": "6p1S8B8wA70",
"url": "https://www.youtube.com/watch?v=6p1S8B8wA70",
"title": "Trump prend une raclée, Poutine est inquiet",
"status": "ok",
"language": "fr",
"originalLanguage": "fr",
"isOriginalLanguage": true,
"isAutoGenerated": true,
"availableLanguages": [{"code": "fr", "autoGenerated": true}],
"translatedTo": null,
"segments": [{"start": 0, "end": 3.04, "text": "Les élections de mi-mandat aux"}],
"text": "Les élections de mi-mandat aux ...",
"segmentCount": 460,
"note": null,
"source": "input"
}

Run it from your code

Python:

from apify_client import ApifyClient
client = ApifyClient("<YOUR_APIFY_TOKEN>")
run = client.actor("alexmorain/youtube-transcript-scraper").call(run_input={
"videos": ["https://www.youtube.com/watch?v=UF8uR6Z6KLc"],
"translateTo": "en",
})
for row in client.dataset(run["defaultDatasetId"]).iterate_items():
if row["type"] == "transcript" and row["status"] == "ok":
print(row["title"], row["segmentCount"])

JavaScript:

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: '<YOUR_APIFY_TOKEN>' });
const run = await client.actor('alexmorain/youtube-transcript-scraper').call({
channels: ['@stanford'],
maxVideosPerSource: 20,
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items.filter((r) => r.status === 'ok').length, 'transcripts');

cURL:

curl -X POST "https://api.apify.com/v2/acts/alexmorain~youtube-transcript-scraper/run-sync-get-dataset-items?token=<YOUR_APIFY_TOKEN>" \
-H "Content-Type: application/json" \
-d '{"videos": ["UF8uR6Z6KLc"], "language": "original"}'

Watch a channel for new videos

Turn on Watch mode, list your channels and schedule the actor (daily, for example). Each run returns only the videos it never delivered before, for each channel, and sends the summary to your webhook. The memory lives in a named key-value store in your own account. A video that ended in timeout or error is tried again next time; one with no captions is not, because YouTube rarely adds them later.

Tips

  • When a video has both captions uploaded by the creator and YouTube's speech recognition in the same language, the creator's captions win.
  • An unknown language code in language or translateTo stops the run before anything is charged, with a message naming the code.
  • Leave language on original unless you need a given track. With a code, a video that lacks that track returns its original language instead, and note says so.
  • Use translateTo rather than language when you want every row in one language: translation works from the original track, so it covers far more videos.
  • Need plain text for an LLM? Read text and ignore segments.
  • Long videos cost the same as short ones: a 2-hour talk is one transcript.

FAQ

How do I get the transcript of a YouTube video in another language?

Set Translate to to the language code you want, for example en or es. YouTube's automatic translation is applied to the original track. The price does not change.

Why do other transcript tools return nothing for some non-English videos?

Many tools ask for an English track by default. On a video that only has French or Spanish captions, that request comes back empty, the same answer as a video without captions. This actor reads the list of tracks first and asks for one that exists.

What is an auto-dubbed video and why does it matter?

YouTube can add an AI-dubbed audio track, often in English, to a video recorded in another language. The default captions then follow the dub. This actor detects the track YouTube marks as original and returns it, so you get what the creator said, not a machine dub of it.

Does it work on videos without captions?

No. If a video has no caption track at all, not even automatic ones, there is no transcript to return. The row comes back with status: no_transcript and is not charged. Music videos, very short clips and some live replays often fall in this case.

Can I get transcripts of a whole YouTube channel?

Yes. Put the channel URL or @handle in Channels and raise Max videos per channel or search (up to 500). Shorts are not included; the actor reads the Videos tab.

Does it handle YouTube Shorts?

A Short given by URL in Videos works like any video if it has captions. Channel listing skips Shorts.

Are auto-generated captions included?

Yes. Most videos only have YouTube's speech recognition captions. isAutoGenerated tells you which kind you got.

How fast is it?

Transcripts are produced in batches by the data provider. Most batches are ready in one to two minutes, whatever their size, and a few take up to five. A run of a few videos takes about as long as a run of a few hundred: on our tests, 500 transcripts came back in under ten minutes, 200 transcripts with 20 comments each in under seven.

The actor reads public videos only, without logging in. Transcripts can be protected by copyright: check how you plan to use them, especially for republishing. For personal data in comments, you are responsible for complying with GDPR and similar laws.

Something is missing or broken

Open an issue on the actor page with the video id and the run id. Rows that fail show a status and a note that help find the cause.

Guide and use cases

Use cases for outbound and automation, the output fields, pricing and common questions are on the YouTube Transcript Scraper page of my site. The Social media and content family page puts it side by side with the other scrapers of the same family.