YouTube Subtitles Downloader & Transcript - Video to Text
Pricing
from $0.70 / 1,000 transcripts
YouTube Subtitles Downloader & Transcript - Video to Text
Returns videoId, language, plain‑text transcript, timestamps and optional auto‑translated text for each requested video, playlist or channel. Perfect for content creators, researchers and SEO analysts who need bulk YouTube subtitles without an API key.
Pricing
from $0.70 / 1,000 transcripts
Rating
0.0
(0)
Developer
Martin Petrov
Maintained by CommunityActor stats
0
Bookmarked
1
Total users
1
Monthly active users
6 hours ago
Last modified
Categories
Features
Share
YouTube Transcript & Subtitles Scraper - Bulk
Extract YouTube transcripts, subtitles and captions in bulk from single videos, whole playlists and entire channels. Works as a simple YouTube transcript API: no YouTube API key and no login needed. Get clean plain text, timed segments, sentence-level text or readable timestamped text, in any language YouTube offers, with optional automatic translation.
Built for reliability: every video is retried with a fresh proxy session, no hidden proxy upgrades, and videos without subtitles are never charged.
What can it do?
- Videos, playlists and channels in one run. Paste any mix of video URLs, video IDs, playlist URLs and channel URLs or
@handles. Playlists and channels are expanded into their videos. - Choose your language. Give a list of preferred languages (
en,de,es…). The first one available is used, with an optional fallback to any available language. - Human captions first. Creator-written captions are preferred over auto-generated ones when both exist.
- Translate on the fly. Set
translateToand every transcript is returned in that language using YouTube's own translation. - Five output formats. Full, plain text, timed segments, sentences, or timestamped text ready for LLMs, RAG pipelines, summaries and subtitles.
- Pay only for results. Videos with disabled or missing subtitles are skipped and listed in the
RUN_SUMMARYrecord, so you know exactly why.
Use cases
- Feed video content into ChatGPT, Claude or a RAG pipeline
- Summarize podcasts, lectures, webinars and interviews
- Content research and SEO: mine what top channels talk about
- Market and competitor research across a whole channel
- Build datasets for NLP and AI training
- Repurpose videos into blog posts, newsletters and social posts
Input
| Field | Type | Description |
|---|---|---|
startUrls | array | Video URLs or IDs, playlist URLs, channel URLs or @handles, one per line. Required. |
languages | array | Preferred language codes in order, e.g. ["en", "de"]. Default ["en"]. |
preferManualCaptions | boolean | Prefer creator-written captions over auto-generated. Default true. |
fallbackToAnyLanguage | boolean | Use any available language if none of the preferred ones exists. Default true. |
translateTo | string | Translate every transcript into this language code. Optional. |
outputFormat | string | full, text, segments, sentences or timestamped. Default full. |
maxVideosPerSource | integer | Newest videos taken from each playlist or channel. Default 50. |
maxItems | integer | Hard limit on transcripts saved (and paid for). Default 1000. |
maxConcurrency | integer | Videos processed in parallel (1–10). Default 5. |
proxyConfiguration | object | Apify Proxy settings. Default: Apify Proxy. |
residentialFallback | boolean | Retired (ignored). To use residential proxy, choose it in proxyConfiguration. |
Example input
{"startUrls": ["https://www.youtube.com/watch?v=iG9CE55wbtY","@TED"],"languages": ["en"],"outputFormat": "full","maxVideosPerSource": 25,"maxItems": 100}
Output
One dataset row per video:
{"videoId": "iG9CE55wbtY","url": "https://www.youtube.com/watch?v=iG9CE55wbtY","title": "Example talk title","channelName": "TED","channelUrl": "https://www.youtube.com/@TED","language": "English","languageCode": "en","isAutoGenerated": false,"translatedTo": null,"wordCount": 3120,"transcriptDurationSeconds": 1150.4,"text": "Good morning. How are you? It's been great, hasn't it? ...","segments": [{ "start": 15.04, "duration": 3.2, "text": "Good morning. How are you?" }],"sentences": [{ "start": 15.04, "text": "Good morning." }],"scrapedAt": "2026-10-26T14:00:00Z"}
Values above are illustrative.
Export as JSON, CSV, Excel or HTML, or read results through the Apify API.
Pricing
Pay per result: you are charged only for transcripts saved to the dataset. Videos that have no usable subtitles cost nothing. Set maxItems to cap the cost of a run.
Monitor channels: only new videos every day
Turn on onlyNew and schedule the actor (Schedules -> daily). Each run returns, and bills, only videos that got no transcript in an earlier run with the same monitorId. Videos whose captions are not ready yet are tried again next run.
{"startUrls": ["https://www.youtube.com/@lexfridman", "https://www.youtube.com/@hubermanlab"],"maxVideosPerSource": 10,"outputFormat": "text","onlyNew": true,"monitorId": "podcasts-daily"}
Call it from your code (Apify API):
curl -X POST "https://api.apify.com/v2/acts/martin10291~youtube-transcript-bulk-extractor/run-sync-get-dataset-items?token=YOUR_APIFY_TOKEN" \-H "Content-Type: application/json" \-d '{"startUrls":["https://www.youtube.com/@lexfridman"],"maxVideosPerSource":5,"outputFormat":"text","onlyNew":true,"monitorId":"lex"}'
The RUN_SUMMARY record shows alreadyDone (skipped as old), skipReasons and lastError.
Tips
- For channels, use
maxVideosPerSourceto take only the newest videos. - Use
outputFormat: "text"for the smallest output when you only need plain text for an LLM. - If YouTube blocks many videos on a large run, pick a residential group in
proxyConfigurationyourself. Blocked videos are listed in RUN_SUMMARY and never charged. - Check the
RUN_SUMMARYrecord in the key-value store to see every skipped video and the reason.
FAQ
Does it work without subtitles? No. The actor reads YouTube's own captions (manual or auto-generated). If a video has subtitles disabled, it is skipped and not charged.
Can I get transcripts in another language? Yes. Use languages to pick from available tracks, or translateTo to translate any transcript.
Do I need a YouTube account or API key? No. The actor uses only publicly available caption data and does not log in.
How many videos can I process? Thousands per run. Use maxItems and maxVideosPerSource to control the size.
More Actors by Martin Petrov
Pair with YouTube Channel Monitor to get every new video from the channels you follow, then pull its transcript automatically.
- YouTube Channel Monitor (New Videos & Shorts): Only new videos, Shorts and streams from many channels, with exact date, views and likes.
- Google Trends Scraper: Interest over time, regions and related queries for many keywords, compared on one scale.
- Remote Jobs Scraper (8 Job Boards): Remote jobs from RemoteOK, Remotive, We Work Remotely, HN Who is Hiring and more, deduplicated.
- ATS Jobs Scraper (Greenhouse, Lever, Ashby): Company job boards from 5 ATS platforms with salaries, filters and only-new mode.
- Reddit Scraper (Posts, Comments & Search): Posts and comments from subreddits, search and users, with an only-new mode for brand monitoring.
- Local Business Leads Finder (Emails, Phones): Local businesses by category and city with company emails, phones and websites, MX-checked.
Legal
This actor extracts publicly available caption data only. It does not log in, bypass paywalls or access private content. You are responsible for using the data in line with YouTube's Terms of Service, copyright law and applicable regulations such as GDPR.
Last updated: October 2026.
Use it as an integration (chain it after another actor)
This actor can run automatically after any actor whose results contain a YouTube link.
- Open the source actor in Apify Console and go to the Integrations tab.
- Click Add integration -> Actor, pick this actor and set Dataset to
{{resource.defaultDatasetId}}(leavestartUrlsempty). Set your other options and save. - When the source run finishes, this actor reads its dataset and processes every row. The field is detected automatically; set Dataset field if your column has another name.
You can also start it by hand with datasetId (pick a dataset or paste its ID) to process an existing dataset.
Good sources to chain from:
- streamers/youtube-scraper (YouTube Scraper)
- martin10291/youtube-channel-monitor
- martin10291/youtube-video-lookup
- martin10291/youtube-shorts-clip-finder