youtube transcript fetcher

YouTube Transcript Extractor

This actor extracts the full transcript text from a YouTube video.

Given a YouTube video URL, it automatically opens the video’s transcript panel and collects all spoken segments into a single plain-text transcript, which is saved to the actor’s dataset.

This is useful for summarization, search indexing, AI analysis, or archiving spoken content from YouTube videos.

⸻

How it works

For a given YouTube video URL, the actor: 1. Opens the video page 2. Expands the video description if needed 3. Finds and clicks the “Show transcript” button 4. Waits for the transcript panel to load 5. Extracts all transcript segments 6. Combines them into one text output 7. Saves the transcript to the dataset

⸻

Input

{ "startUrls": [ { "url": "https://www.youtube.com/watch?v=VIDEO_ID" } ], "proxyConfig": {} }

Input fields • startUrls – List of YouTube video URLs to process (only the first URL is used) • proxyConfig – Optional proxy configuration

⸻

Output

The actor stores results in the default dataset.

Each dataset item contains:

{ "transcript": "Full transcript text from the video..." }

•	Transcript text is returned as a single combined string
•	Timestamps are not included

⸻

Notes & limitations • Only works for videos where YouTube transcripts are available • Processes one video per run • Does not include timestamps or speaker labels • Relies on YouTube’s current interface, which may change over time