YouTube Transcript Engine — Subtitles & Captions API avatar

YouTube Transcript Engine — Subtitles & Captions API

Pricing

$1.00 / 1,000 results

Go to Apify Store
YouTube Transcript Engine — Subtitles & Captions API

YouTube Transcript Engine — Subtitles & Captions API

Batch YouTube transcripts & subtitles — video URLs, whole channels, keyword search, channel watch. Multi-client fallback + rotating proxy retry for reliability. $0.001/video.

Pricing

$1.00 / 1,000 results

Rating

0.0

(0)

Developer

William Billings

William Billings

Maintained by Community

Actor stats

0

Bookmarked

2

Total users

1

Monthly active users

5 days ago

Last modified

Features

Limited permissions

Share

Transcripts at pipeline scale — batch video URLs, entire channels, or keyword search → clean timestamped transcript rows. Built for AI workloads: RAG ingestion, LLM fine-tuning, summarization, media monitoring. No API key needed.

Who it's for

  • AI/LLM builders — bulk transcript ingestion for RAG, training data, and summarization pipelines
  • Media monitoring — watch channels, auto-transcript every new upload on a schedule
  • Researchers & analysts — searchable text of talks, interviews, briefings at scale
  • Content teams — transcripts for repurposing, SEO, and accessibility

Why this one

YouTube gates captions per client and the gates move — single-client scrapers break weekly (that's why the biggest incumbents run double-digit failure rates). This engine fails over across 4 client identities (Android → TV → iOS → embedded) until captions resolve — and when YouTube bot-gates the datacenter IP itself, it retries through your proxy config (optional, residential recommended for heavy pipelines). Direct connections stay free; proxies only kick in when needed.

Modes

ModeWhat you get
transcriptBatch list of video URLs → transcript rows (one row per video)
channelLatest N videos from a channel → all transcripts
searchKeyword → top videos + their transcripts
watchChannel list → newest videos auto-transcribed — schedule it

Output

One row per video: videoId, url, title, channelName, language, autoGenerated, segmentCount, text (full), segments[] (timestamped), availableLanguages, clientUsed

{
"recordType": "transcript", "videoId": "dQw4w9WgXcQ",
"title": "Rick Astley - Never Gonna Give You Up",
"language": "en", "autoGenerated": true, "segmentCount": 210,
"text": "We're no strangers to love ...",
"segments": [{"startMs": 1360, "durationMs": 1680, "text": "..."}]
}

Example inputs

{ "mode": "transcript", "videoUrls": ["https://youtu.be/abc123...", "dQw4w9WgXcQ"] }
{ "mode": "channel", "channelUrls": ["@TED"], "maxVideos": 20 }
{ "mode": "search", "query": "LLM agents tutorial", "maxVideos": 10 }
{ "mode": "watch", "channelUrls": ["@TED", "@lexfridman"], "language": "en" }

Notes

  • Handles manual captions + auto-generated (ASR), all available languages listed per video
  • Videos without captions emit no_transcript rows with the reason — pipelines stay clean
  • Accepts URLs, bare video IDs, /shorts/ links, @handles, and UC... channel IDs

Pricing

$0.001 per video row — no subscription, no key.

More from this developer