Catalyst Labs / Guides

YouTube Transcript API for AI Agents (MCP, JSON, SRT)

AI agents often need the words spoken in a video: to summarise it, research a competitor channel or answer questions. Our YouTube Transcript API on Apify returns transcripts with timestamps plus video metadata as JSON, SRT, VTT, plain text or Markdown.

Option 1: connect it to an agent over MCP

Add this MCP server URL to Claude, Cursor or any MCP client (Apify token required):

https://mcp.apify.com/?tools=catalyst-labs/youtube-transcript-api

The agent gets one tool that takes video, Shorts, playlist, channel or search inputs.

Option 2: call the REST API

curl -X POST "https://api.apify.com/v2/acts/catalyst-labs~youtube-transcript-api/run-sync-get-dataset-items?token=$APIFY_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"videos":["https://www.youtube.com/watch?v=jNQXAC9IVRw"],"languages":["en"],"outputFormats":["text","segments"]}'

videos also takes Shorts, playlist and channel URLs; use searchQueries for YouTube search. Set "mode":"listCaptionTracks" to check available caption languages for free. Full options are on the Store page input tab.

What you get

• Timestamped segments plus full text, with auto-generated vs human captions flagged
• Language choice and a free check of which caption languages exist
• Channel monitoring that only returns new videos, with date filters
• You pay only for transcripts delivered: $0.005 per run + $0.008 per transcript on the Free plan, lower on higher tiers

Limits

Videos with no captions return no transcript (and aren't charged).

Try it on the Apify Store