YouTube Transcript Scraper - Subtitles & Captions API

What does YouTube Transcript Scraper do?

YouTube Transcript Scraper gets the transcripts of YouTube videos and Shorts, one by one or a whole channel, playlist or search at a time: the full text, timestamped segments like YouTube's Show transcript panel, and SRT or WebVTT subtitle files, in the video's own language or in every language the video has. It works as an unofficial YouTube transcript API: no API key, no quota and no login.

Paste video links, channels or playlists, or type search terms, pick a subtitle language, and download the results as JSON, CSV or Excel, or get them through the API. Subtitles from the creator are used first, and YouTube's auto-generated captions fill in when there are none.

Why use YouTube Transcript Scraper?

• 📺 Whole channels, playlists and searches: add a channel to get its newest videos, Shorts or live replays, a playlist for every video in it, or a search term for the top results. You choose how many videos to take from each.
• 🌍 Every language in one run: pick one language, or choose All languages to get every subtitle track a video has, including the captions of YouTube's AI-dubbed audio tracks.
• 🎯 The right language: the same list as Subtitles/CC in the YouTube player. Subtitles written by the creator come before auto-generated captions, and each result says which one you got.
• ⏱️ Timestamps your way: plain text for AI and search, timestamped segments with start and end times in seconds, and optional SRT / WebVTT files.
• 📊 One line per row: the Segments view lists every timestamped line as its own row, so the CSV or Excel export is ready to use.
• 🧾 Clear skips, no charge: videos without the language you asked for, without captions, or that need a sign-in are skipped, not charged, and listed with the reason.
• 🎬 Videos, Shorts and live replays: paste any video link, or just the video ID. Exact view count, duration, publish time and channel come with every transcript.
• ⚡ Fast: 20 videos at a time. In our tests, 60 videos took about 10 seconds.

Popular use cases: feeding transcripts to ChatGPT, Claude or other AI tools, RAG and search indexes, video summaries and show notes, content research and SEO, translation and subtitling work, accessibility, and building datasets.

	YouTube Transcript Scraper	YouTube Data API	Copying by hand
Transcripts of any public video	Yes	Only your own videos (OAuth)	Yes, slowly
API key / Google Cloud project	Not needed	Required	Not needed
All languages in one go	Yes	One request per track	One at a time
Timestamps and SRT / VTT	Yes	Yes	Timestamps only
Export to CSV / Excel / JSON	Yes	Needs code	Manual

What data can you extract?

Each result is one video in one subtitle language.

Field	Description	Example
transcript	The whole transcript as plain text	This is a 3. It's sloppily written and rendered...
segments	Timestamped lines: start and end in seconds, start_time as shown on YouTube, text	{"start": 4.22, "end": 5.4, "start_time": "0:04", "text": "This is a 3."}
language	Subtitle name as shown in the Subtitles/CC menu	English, English (auto-generated)
language_code	Language code	en
is_auto_generated	true for captions YouTube made from the speech, false for subtitles from the creator	false
available_languages	Every subtitle track the video has	[{"language_code": "ar", "language": "Arabic", ...}]
subtitles	The whole SRT or WebVTT file as text, when Subtitle file is set	1\n00:00:04,220 --> 00:00:05,400\nThis is a 3.
title	Video title	But what is a neural network? | Deep learning chapter 1
channel_name, channel_id, channel_url	Channel name, ID and link	3Blue1Brown
published_at	Exact publish time (UTC)	2017-10-05T15:11:25Z
video_duration, duration_seconds	Video length	18:40, 1120
view_count	Exact view count	24649096
video_url, video_id	Video link from your input and video ID	aircAruvnKk
description, thumbnail_url	Video description and thumbnail
source, source_position	The link or search term the video came from, and its position there (newest first for channels)	@3blue1brown, 1

How to scrape YouTube transcripts

1. Click Try for free on this page and sign in to Apify (a free account is enough).
2. Fill in one or more options: Videos (video or Shorts links), Channels (for example @3blue1brown), Playlists (playlist links) or Search YouTube (search terms).
3. For channels, playlists and searches, set how many videos to take from each. Then pick a Subtitle language: Original language, All languages, or a specific language.
4. Click Start. The log shows each video as it's done.
5. Download the transcripts as JSON, CSV, Excel or HTML, or get them through the API. Use the Segments view for one timestamped line per row.

How much does it cost to scrape YouTube transcripts?

You only pay for the transcripts you get: one video in one language is one result. With All languages, each language counts as a result. Listing the videos of a channel, playlist or search is free, and videos that are skipped (no captions, no subtitles in your language, age-restricted, members-only or unavailable) are not charged. See the Pricing tab for the current price.

To keep a run within budget, set a Max cost per run in the run options. The Actor stops as soon as the limit is reached.

New to Apify? The free plan includes monthly platform credit, so you can try this Actor at no cost.

Input

Field	Type	Default	Description
Option 1 — Videos			One transcript per video.
YouTube video links (urls)	List	—	Video, Shorts or live links, or video IDs, one per line. Example: https://youtu.be/aircAruvnKk
Option 2 — Channels			The newest videos of each channel.
Channels (channels)	List	—	Channel handle, channel link or channel ID, one per line. Example: @3blue1brown
Channel tab (channel_tab)	Select	videos	videos, shorts or streams (Live), like the tabs on the channel page. A channel link ending in /shorts or /streams uses that tab.
Max videos per channel (max_videos_per_channel)	Number	20	How many of the newest videos to take from each channel. 0 takes all.
Option 3 — Playlists			The videos of each playlist, in playlist order.
Playlists (playlists)	List	—	Playlist links or IDs, one per line. Example: https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi
Max videos per playlist (max_videos_per_playlist)	Number	20	How many videos to take from each playlist, from the top. 0 takes the whole playlist.
Option 4 — Search YouTube			The top videos for a search.
Search terms or search result links (search_keywords)	List	—	Search terms (videos only, like YouTube's Type: Video filter), or YouTube search results page links to keep your filters. Example: neural networks explained
Max videos per search (max_videos_per_search)	Number	20	How many of the top videos to take from each search.
Transcript options			These apply to all four options.
Subtitle language (language)	Select	original	original (the language spoken in the video, the one Show transcript opens), all (every subtitle track), or a language code such as es, pt, ja or zh-Hans. Same list as Settings → Subtitles/CC in the YouTube player.
Include auto-generated captions (include_auto_generated)	Checkbox	true	Use YouTube's auto-generated captions when there are no subtitles from the creator. Turn off for creator subtitles only.
Subtitle file (subtitle_format)	Select	none	srt or vtt adds the whole subtitle file as text in subtitles.

Transcripts of a few videos in their own language

{
    "urls": [
        "https://www.youtube.com/watch?v=aircAruvnKk",
        "https://www.youtube.com/shorts/CEJXqm2eiJ0"
    ]
}

Spanish subtitles, from the creator only, as SRT files

{
    "urls": ["https://www.youtube.com/watch?v=aircAruvnKk"],
    "language": "es",
    "include_auto_generated": false,
    "subtitle_format": "srt"
}

The 20 newest videos of a channel

{
    "channels": ["@3blue1brown"],
    "max_videos_per_channel": 20
}

The 10 newest Shorts of a channel

{
    "channels": ["@LinusTechTips"],
    "channel_tab": "shorts",
    "max_videos_per_channel": 10
}

A whole playlist, plus the top 10 videos for a search

{
    "playlists": ["https://www.youtube.com/playlist?list=PLZHQObOWTQDNU6R1_67000Dx_ZCJB-3pi"],
    "max_videos_per_playlist": 0,
    "search_keywords": ["neural networks explained"],
    "max_videos_per_search": 10
}

Every subtitle language of a video

{
    "urls": ["https://www.youtube.com/watch?v=y1hZ3EalPQU"],
    "language": "all"
}

Output

Each transcript is one item in the dataset, in the order of your input: videos first, then channels, playlists and searches. Here is a real result (text, segments and lists shortened):

{
    "video_url": "https://www.youtube.com/watch?v=aircAruvnKk",
    "video_id": "aircAruvnKk",
    "title": "But what is a neural network? | Deep learning chapter 1",
    "channel_name": "3Blue1Brown",
    "channel_id": "UCYO_jab_esuFRV4b17AJtAw",
    "channel_url": "https://www.youtube.com/channel/UCYO_jab_esuFRV4b17AJtAw",
    "published_at": "2017-10-05T15:11:25Z",
    "video_duration": "18:40",
    "duration_seconds": 1120,
    "view_count": 24649096,
    "language": "English",
    "language_code": "en",
    "is_auto_generated": false,
    "transcript": "This is a 3. It's sloppily written and rendered at an extremely low resolution of 28x28 pixels, but your brain has no trouble recognizing it as a 3. And I want ...",
    "segments": [
        {"start": 4.22, "end": 5.4, "start_time": "0:04", "text": "This is a 3."},
        {"start": 6.06, "end": 10.713, "start_time": "0:06", "text": "It's sloppily written and rendered at an extremely low resolution of 28x28 pixels,"}
    ],
    "available_languages": [
        {"language_code": "ar", "language": "Arabic", "is_auto_generated": false},
        {"language_code": "en", "language": "English", "is_auto_generated": false},
        {"language_code": "en", "language": "English (auto-generated)", "is_auto_generated": true}
    ],                                    // 31 tracks for this video
    "subtitles": null,                    // SRT or WebVTT text when "Subtitle file" is set
    "description": "What are the neurons, why are there layers, and what is the math underlying it? ...",
    "thumbnail_url": "https://i.ytimg.com/vi_webp/aircAruvnKk/sddefault.webp",
    "source": "https://www.youtube.com/watch?v=aircAruvnKk",
    "source_position": null              // Position in the channel, playlist or search the video came from
}

The Segments view in the Output tab shows one timestamped line per row (title, language, time, start, end, text), which is the easiest way to work with transcripts in a spreadsheet.

Tips and limitations

• Original language is the language spoken in the video. If the creator uploaded subtitles in that language you get those, otherwise YouTube's auto-generated captions.
• Language matching: a language also matches its regional versions, so Portuguese finds Portuguese (Brazil) and Spanish finds Spanish (Latin America). Chinese (Simplified) and Chinese (Traditional) are kept apart.
• AI-dubbed videos: when YouTube has dubbed a video into other languages, each dubbed language has its own auto-generated captions, and you can pick them like any other language.
• Auto-translate is not available. YouTube blocks automatic translations for tools like this one, so you get the languages the video really has. available_languages shows them.
• Skipped videos: videos with no captions, live streams that are still on air, age-restricted, members-only and private videos are skipped and not charged. The reasons are in SKIPPED_VIDEOS in the Storage tab.
• Long videos are fine: a 4.5-hour course comes back as one result with almost 3,000 segments.
• Channels: you get the newest videos of the tab you pick in Channel tab (Videos, Shorts or Live). A channel link ending in /shorts or /streams uses that tab instead. Members-only videos are not included, because their captions need a sign-in.
• Playlists: use the playlist link (youtube.com/playlist?list=...). A video link that also has &list= gets just that video. Mix playlists (auto-generated radio) can't be listed.
• Search filters: search on YouTube, set filters such as upload date or duration, and paste the results page link. Your filters are kept.
• Max videos counts videos, not transcripts: videos without subtitles are skipped and not charged, so you can get fewer transcripts than the number you set. The same video found in several places is processed once.
• Pasted a link in the wrong box? It still works: links are recognised by what they point to, so a channel link in the Videos box is treated as a channel.

FAQ

Is it legal to scrape YouTube transcripts?

This Actor only collects data that YouTube shows publicly. Our Actors are ethical and do not extract any private user data, such as email addresses, gender, or location. They only extract what the user has chosen to share publicly. We therefore believe that our Actors, when used for ethical purposes by Apify users, are safe. However, you should be aware that your results could contain personal data. Personal data is protected by the GDPR in the European Union and by other regulations around the world. You should not scrape personal data unless you have a legitimate reason to do so. If you're unsure whether your reason is legitimate, consult your lawyers.

Do I need a YouTube API key or account?

No. Add your videos, channels, playlists or search terms and start the run.

Can I get the transcripts of a whole YouTube channel?

Yes. Add the channel in Channels (for example @3blue1brown) and set Max videos per channel to the number of newest videos you want, or 0 for all of them. Pick Shorts or Live in Channel tab for Shorts or past live streams.

Can it translate transcripts?

No. YouTube's auto-translate can't be downloaded by tools like this one, so the Actor returns the subtitles the video really has. Many popular videos have subtitles from the creator in several languages, and AI-dubbed videos have captions for each dubbed language. You can also send the transcript to an AI tool for translation.

Why was a video skipped?

Because it has no captions, doesn't have the language you picked, is age-restricted, members-only or private, or no longer exists. Channels that don't exist, Mix playlists and searches with no results are skipped too. Skipped videos are not charged. The run's status message counts them by reason, and SKIPPED_VIDEOS in the Storage tab lists each one with the languages it does have.

How is "All languages" charged?

Each language is one result. A video with subtitles in 20 languages gives 20 results.

Can I get one timestamped line per row?

Yes. Open the Segments view in the Output tab, or export the dataset with that view, and every line of every transcript becomes its own row.

Does it work with YouTube Shorts?

Yes. Paste the Shorts link (or its video ID) like any other video.

Can I use it through the API, with n8n or with AI agents?

Yes. You can run it with the Apify API, the Python and JavaScript clients, integrations such as Make, Zapier and n8n, or the Apify MCP server so AI agents can fetch transcripts themselves.

Related Actors

Actor	What it does
YouTube Channel Scraper - All Videos, Views & Dates	Every video, Short and live stream of a channel
YouTube Search Scraper - Videos & Shorts by Keyword	Videos that match a search term
YouTube Video Details Scraper - Views, Likes & Channel Stats	Full details of any list of video links
YouTube Comments Scraper - Replies, Likes & Shorts	Comments and replies of any video

Support

Found a problem or need another field? Open an issue in the Issues tab and we'll get back to you.