Signal Lab/Workflows/YouTube Transcript Scraper

READY-TO-RUN WORKFLOW

Export a YouTube transcript for RAG or research

Extract a public video's full caption text, language details, optional timestamped segments, and useful source metadata without a YouTube API key.

BUILT FORAI builders, researchers, content teams, analysts, and automation developers
LIVE PRICING$0.00005 per run + $0.00001 per successful transcript

THE OUTCOME

A clean transcript row with the video title, channel, language, caption type, duration, views, source URL, and optional timestamped segments.

  1. 1

    Open the Actor and run the prefilled one-video input unchanged.

  2. 2

    Inspect the transcript, language, caption type, and source metadata.

  3. 3

    Replace the sample with a public YouTube URL or 11-character video ID.

  4. 4

    Enable timestamped segments or raise the video cap only after the first bounded result succeeds.

STRUCTURED OUTPUT

Fields ready for export and automation

Video IDSource URLTitleChannelLanguageCaption typeDurationView countTranscriptTimestamped segments

COMMON QUESTIONS

Before you run the workflow

Does it transcribe audio?

No. It extracts caption tracks YouTube exposes publicly; it does not download media or generate speech-to-text.

Do I need a YouTube API key?

No. Add public watch, Shorts, live, embed, youtu.be URLs, or raw 11-character video IDs.

Why can a video fail?

Private, deleted, age- or region-restricted videos and videos without public captions cannot produce transcript rows.

START WITH A BOUNDED SAMPLE

Open the Actor, keep the sample small, and inspect the first dataset.

Run this exact example