Youtube Subtitles Scraper

YouTube Captions & Transcripts Scraper (Apify Actor)

Extract clean, timestamped YouTube subtitles at scale. Perfect for creators, researchers, SEO teams, accessibility workflows, and AI/NLP pipelines that need accurate multi‑language transcripts fast.

Quick start

• Open the Actor on Apify and paste one or many YouTube URLs.
• Select Language mode: “Get all available” or “Get by code” (e.g., en, es).
• (Optional) Keep auto‑generated subtitles enabled for broader coverage.
• Run the Actor. When finished, export the dataset as JSON or CSV.
• Use the results for summaries, translations, search, or content repurposing.

What it does

• Bulk‑extracts captions/transcripts from any public YouTube videos
• Multi‑language: fetch all available languages or only the ones you choose
• Keeps timestamps and ordering for easy quoting and analysis
• Handles auto‑generated subtitles when official ones aren’t available
• Designed for reliability with smart retries and proxy rotation on errors
• Outputs structured data you can export as JSON, CSV, or use directly in apps

Who it’s for

• Content creators and editors repurposing videos into articles, posts, or newsletters
• Researchers and journalists collecting quotes and evidence from video sources
• SEO and marketing teams mining topics, entities, and keywords from video content
• Accessibility and compliance teams generating readable caption text
• AI & data teams feeding high‑quality text into summarization, RAG, or LLM training

Popular use cases

• Summarize long videos into notes, briefs, or articles
• Create highlight reels and pull timestamped quotes for social posts
• Translate and localize scripts to reach global audiences
• Index channel archives for topic discovery and search
• Build clean training datasets for speech and language models

---

Simple workflow

1. Add one or many YouTube URLs (bulk friendly)
2. Choose language preference:
   • Get all available languages, or
   • Specify language codes (e.g., en, es, de)
3. Run and download results from the dataset as JSON/CSV

That’s it — no coding required.

---

Input options (plain English)

• Video URLs: paste one or more links. You can bulk‑edit the list.
• Language mode:
  • Get by code — provide one or more language codes.
  • Get all available — collect every transcript language offered by the video.
• Language codes (when using “Get by code”): examples en, en, es.
• Fetch auto‑generated subtitles: include YouTube’s auto captions when official ones aren’t present.
• Max retries: how many times to retry a transcript fetch if a request temporarily fails.
• Proxy settings: uses Apify Proxy under the hood. On network/proxy errors, the actor rotates to a fresh proxy to keep your run moving.

Notes

• You can add/remove URLs one‑by‑one or paste a whole list.
• URLs are deduplicated while preserving order.

---

Output at a glance

For each video, the actor saves one dataset item with:

• video_id and original url
• supported_languages: a list of language codes found
• transcripts: a mapping of language code → transcript entries (with timestamps)

You can export the dataset to JSON/CSV directly in Apify.

---

Why this tool

• Scale: process dozens or thousands of videos in one run
• Accuracy: preserves timestamps and structure for downstream analysis
• Coverage: fetch every available language or narrow to what you need
• Reliability: on‑error proxy rotation and retries reduce flaky failures
• Ready for AI: clean text that plugs straight into summarization, search, and RAG

---

Tips & best practices

• Exploring a channel? Start with “Get all available” languages to see coverage, then narrow to specific codes.
• Keep auto‑generated subtitles enabled to maximize results on videos without official captions.
• For lighter outputs, specify just the languages you need.
• Always respect platform terms, privacy, and content rights.

---

FAQ

• Does it work on private or members‑only videos?
  • No. It works with publicly available videos that offer captions/transcripts.
• Does it bypass paywalls or protections?
  • No. It only collects data that YouTube exposes for public content.
• Are timestamps preserved?
  • Yes. Each transcript entry includes start times so you can quote precisely.
• Can I run this in bulk?
  • Yes. Paste a list of URLs and the actor will process them in order.