YouTube Transcript Scraper

Get the full transcript of any YouTube video or Short: timestamped segments, clean plain text, or both. Paste in one URL or thousands. You pay only for transcripts actually delivered.

Why this one

• Pay per transcript, nothing else. Videos with no captions, private or deleted videos, bad links and temporary errors are all free. They still appear in your results with a reason, so you know what happened.
• No surprise bills. Maximum videos is a hard cap, and the actor checks your maximum cost per run before every request. It stops cleanly when the budget is used rather than going over.
• Any language, including translations. Ask for a caption language (en, es, pt-BR …). If the video doesn't have it, YouTube's automatic translation is returned where available and flagged with autoTranslated: true. Or ask for the original spoken language only, skipping auto-dubbed tracks.
• AI-ready text. Optionally strip caption annotations such as [music] and [applause] before the text reaches your LLM or RAG pipeline.
• Every link format. watch?v=, youtu.be, Shorts, live replays, embeds, mobile and YouTube Music links, or bare video IDs. Duplicates are removed before anything is fetched.
• Fast. Fetches several videos in parallel, retries transient errors automatically, and serves recently fetched transcripts from cache.

Input

Field	Description	Default
videoUrls	Video or Shorts URLs, or video IDs	required
language	Caption language code; empty means YouTube's default track	empty
originalAudio	Return captions in the original spoken language only (overrides language)	false
outputFormat	both, segments or text	both
removeSoundTags	Strip [music], [applause] and similar annotations	false
maxVideos	Hard cap on videos processed per run	100
cacheMaxAge	Accept a cached transcript up to 1d, 3d, 7d, 14d, 30d, or none for always fresh	7d
includeFailures	Add a free row for each video without a transcript	true
maxConcurrency	Parallel requests (1–20)	5

Example:

{
  "videoUrls": [
    "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
    "https://youtu.be/jNQXAC9IVRw",
    "https://www.youtube.com/shorts/abcdefghijk"
  ],
  "language": "en",
  "outputFormat": "both",
  "maxVideos": 50
}

Output

One row per video.

{
  "status": "ok",
  "videoId": "jNQXAC9IVRw",
  "url": "https://www.youtube.com/watch?v=jNQXAC9IVRw",
  "type": "video",
  "input": "https://youtu.be/jNQXAC9IVRw",
  "language": "English",
  "autoTranslated": false,
  "requestedLanguage": "en",
  "segmentCount": 6,
  "wordCount": 42,
  "durationSeconds": 19,
  "text": "All right, so here we are in front of the ...",
  "segments": [
    { "startMs": 1200, "endMs": 4100, "startTime": "0:01", "text": "All right, so here we are" }
  ],
  "scrapedAt": "2026-10-19T09:40:00.000Z"
}

autoTranslated is true when you asked for a language the video doesn't have and YouTube's machine translation was returned instead. In that case language shows the code you requested.

Videos without a transcript come back like this (free):

{
  "status": "no_captions",
  "videoId": "...",
  "url": "...",
  "errorMessage": "No transcript available. The video may have no captions, or may be private, deleted or unavailable.",
  "charged": false
}

status	Meaning	Charged
ok	Transcript delivered	Yes
no_captions	No transcript available: no captions, or the video is private, deleted or unavailable	No
not_found	Video not found	No
invalid_input	Not a YouTube video link or ID	No
error	Temporary failure after retries; safe to re-run	No

The Timestamped segments view in the Output tab gives one row per segment, ready to export as CSV for subtitles or analysis. A run summary is saved to the key-value store as SUMMARY.

Use cases

• Feed video content into LLMs, RAG pipelines and summarisers
• Repurpose videos into blogs, newsletters and social posts
• Search, quote and timestamp-link spoken content
• Research competitors, creators and topics at scale
• Build subtitle files and accessibility tooling

Pricing

Pay per event: you're charged once per transcript delivered (status: "ok"), plus Apify's small per-run start fee. Nothing else. Set Maximum cost per run in the run options and the actor will never exceed it.

Limits

• Only videos that have captions (manual or auto-generated) can return a transcript.
• Age-restricted, members-only and private videos usually have no accessible transcript; they are returned as free failures.
• Accepts individual videos. Channel and playlist URLs aren't expanded; pass video URLs instead.

Integrations

Works with the Apify API, Make, Zapier, n8n, LangChain and the Apify MCP server. Call it from your own code:

import { ApifyClient } from 'apify-client';
const client = new ApifyClient({ token: process.env.APIFY_TOKEN });
const run = await client.actor('datara/youtube-transcript-scraper').call({
  videoUrls: ['https://www.youtube.com/watch?v=dQw4w9WgXcQ'],
});
const { items } = await client.dataset(run.defaultDatasetId).listItems();