Video Transcript Scraper: YouTube, TikTok, Instagram & Facebook

Video Transcript Scraper

Turn YouTube, TikTok, Instagram Reels and Facebook videos into text with timestamps, in one run. Paste video URLs and get full transcripts plus ready-to-use SRT and VTT subtitles. It uses the platform's captions when they exist and AI speech-to-text when they don't. No API key, no login.

Supports: YouTube (videos, Shorts, live, embeds) · TikTok · Instagram Reels and video posts · Facebook videos and Reels

---

What does Video Transcript Scraper do?

Video Transcript Scraper turns public videos into structured text with timestamps. Give it a list of video URLs from any supported platform and it returns the full transcript, ready-to-use SRT and WebVTT subtitle documents, individual timed segments, word count, and video metadata (title, author, thumbnail).

For multilingual YouTube workflows, set preferredLanguage to an exact YouTube caption language code such as es, de, or pt-BR. The Actor prefers a manual track and then an auto-generated track in that language. When this option is omitted, selection remains English-first and then falls back to the first available track.

YouTube transcripts come from the video's caption tracks. Instagram and Facebook have no public caption files, so the Actor downloads the public video and transcribes its speech with AI (Whisper). TikTok uses its caption file when the page exposes one and speech-to-text otherwise. The transcriptSource field tells you which one each row used. Speech-to-text finds speech only: a music-only clip returns an error row and costs nothing.

Try it free on the Apify Store.

---

Who is it for?

📢 Content marketers and social media managers

• Repurpose video content into blog posts, newsletters, or social captions
• Extract key quotes from interview-style Reels for ad copy
• Audit influencer content for brand mentions before partnerships

🎓 Researchers and academics

• Analyze linguistic patterns across thousands of TikTok or YouTube videos
• Build datasets of spoken language from social media for NLP training
• Study how topics are discussed across different video platforms

🤖 AI and automation engineers

• Feed transcripts into LLMs for summarization, sentiment analysis, or Q&A
• Build RAG pipelines indexing video content as searchable text
• Automate transcript generation for video archives

💼 SEO and content strategists

• Convert YouTube video content into text for programmatic SEO pages
• Identify top-performing topics by analyzing what high-view videos say
• Extract transcript data from competitor channels for content gap analysis

---

Why use Video Transcript Scraper?

• 🎯 Multi-platform in one run — mix TikTok, YouTube, Instagram, and Facebook URLs in a single input list
• 🎙️ Works without captions — AI speech-to-text covers Instagram, Facebook and TikTok videos that have no caption file
• ⚡ Fast YouTube — one to a few seconds per video, no browser
• 🗣️ Optional YouTube caption language — use preferredLanguage for an exact available track, or omit it for English-first selection
• 🕐 Ready-to-use subtitles — each successful result includes complete srt and vtt documents generated from the timed segments
• 📄 Clean text output — transcriptText is a single clean string, while every segment retains startTime and endTime in seconds
• 🔁 Graceful error handling — failed or no-caption videos don't crash the run; they return an error field
• 🛡️ No proxy setup — the Actor picks the route each platform needs, and proxy traffic is included in the price
• 💳 Pay for delivery — the start fee and the per-video price apply only when a transcript is delivered
• 🔗 No login required — works on public videos without any account credentials
• 📊 Structured JSON output — every result includes platform, videoId, author, thumbnail, duration, language, segments, and word count

---

What data can you extract?

Field	Description	Example
videoUrl	Original input URL	https://youtube.com/watch?v=...
platform	Platform name	youtube / tiktok / instagram / facebook
videoId	Platform video ID	dQw4w9WgXcQ
title	Video title	"10 Python Tips You Didn't Know"
author	Channel or username	"@techcreator"
thumbnailUrl	Thumbnail image URL	https://img.youtube.com/vi/.../maxresdefault.jpg
duration	Duration in seconds	243
description	Video description or post caption, when exposed	"How we built it"
viewCount	View or play count, when exposed	18800000
language	Caption language code, or the language detected by speech-to-text	en
transcriptSource	captions or speech-to-text	speech-to-text
billedMinutes	Started minutes of audio charged on a speech-to-text row, otherwise null	2
transcriptText	Full transcript as plain text	"Today we're going to talk about..."
srt	Complete SubRip subtitle document (null on errors)	1\n00:00:01,250 --> ...
vtt	Complete WebVTT subtitle document (null on errors)	WEBVTT\n\n00:00:01.250 --> ...
segments	Array of timed caption segments	[{startTime, endTime, text}, ...]
segmentCount	Number of caption segments	87
wordCount	Approximate word count	1243
scrapedAt	Extraction timestamp	2026-04-04T12:00:00.000Z
error	Why there is no transcript (null if successful)	"No speech was detected..."

Segment object fields:

Field	Type	Description
startTime	number	Segment start time in seconds
endTime	number	Segment end time in seconds
text	string	Caption text for this segment

---

How much does it cost to extract video transcripts?

This Actor uses pay-per-event pricing: you pay for transcripts delivered, nothing else.

Event	BRONZE price	Charged
Run started	$0.003	Once per run, with the first transcript delivered
YouTube transcript	$0.003	Per video
Instagram transcript	$0.0065	Per started minute of audio
TikTok transcript	$0.012	Per started minute of audio. A TikTok with its own caption file counts as one unit at any length
Facebook transcript	$0.0065	Per started minute of audio

Why per minute? Speech-to-text takes compute in proportion to the audio's length, up to the 10-minute limit. A 40-second Reel is one unit, a 3-minute video is three, and a 10-minute video is ten. Each row's billedMinutes field shows what was charged.

Plan discounts apply to the transcript prices, and the start fee stays the same:

Apify plan	YouTube per video	Instagram and Facebook per minute	TikTok per minute
FREE and BRONZE	$0.003	$0.0065	$0.012
SILVER	$0.0027	$0.00585	$0.0108
GOLD and above	$0.0024	$0.0052	$0.0096

Failed extractions are not charged — a video without captions or speech, a private video or a failed extraction costs nothing. A run that delivers no transcript has no start fee either. A spending limit is respected: the run stops before a transcript it could not pay for.

Real-world cost examples (BRONZE):

• 50 YouTube transcripts: $0.003 + (50 × $0.003) = $0.153
• 50 Instagram Reels under a minute: $0.003 + (50 × $0.0065) = $0.328
• One 10-minute Facebook video: $0.003 + (10 × $0.0065) = $0.068

How to extract video transcripts

1. Go to Video Transcript Scraper on Apify Store
2. Click Try for free
3. In the Video URLs field, paste one or more video URLs (one per line)
4. (Optional) Set Preferred YouTube caption language to an exact code such as es; leave it empty for English-first selection
5. (Optional) Adjust Max retries for flaky connections
6. Click Start and wait for results
7. Download results as JSON, CSV, or Excel — or connect to downstream tools

Supported URL formats:

# YouTube
https://www.youtube.com/watch?v=dQw4w9WgXcQ
https://youtu.be/dQw4w9WgXcQ
https://www.youtube.com/shorts/dQw4w9WgXcQ

# TikTok
https://www.tiktok.com/@username/video/7123456789012345678

# Instagram
https://www.instagram.com/reel/ABC123def/

# Facebook
https://www.facebook.com/reel/123456789

Example input (JSON):

{
  "videoUrls": [
    "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
    "https://www.youtube.com/shorts/5-WE64bwEMI",
    "https://www.tiktok.com/@nasa/video/7234567890123456789"
  ],
  "preferredLanguage": "es",
  "maxRetries": 2
}

Example input — batch content analysis:

{
  "videoUrls": [
    "https://www.youtube.com/watch?v=VIDEO_ID_1",
    "https://www.youtube.com/watch?v=VIDEO_ID_2",
    "https://www.tiktok.com/@creator/video/TIKTOK_ID"
  ],
  "maxRetries": 3
}

---

Input parameters

Parameter	Type	Required	Default	Description
videoUrls	array of strings	✅ Yes	—	Video URLs to transcribe. Mix platforms freely.
preferredLanguage	string	No	English-first	Exact YouTube caption language code (for example es or pt-BR). Manual tracks are preferred over generated tracks. You get an error row if the requested language is unavailable. This option does not affect other platforms or translate captions.
maxVideoDurationSeconds	integer	No	600	Videos needing speech-to-text beyond this length return an error row, not charged (10–600).
maxRetries	integer	No	2	Max retry attempts per URL on failure (0–5).
proxy	object	No	Automatic	Optional proxy override. The default route works for every platform.

Memory: the default of 1024 MB suits every platform. Speech-to-text needs at least 1024 MB.

---

Output examples

Successful YouTube transcript:

{
  "videoUrl": "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
  "platform": "youtube",
  "videoId": "dQw4w9WgXcQ",
  "title": "Rick Astley - Never Gonna Give You Up (Official Music Video)",
  "author": "Rick Astley",
  "thumbnailUrl": "https://i.ytimg.com/vi/dQw4w9WgXcQ/maxresdefault.jpg",
  "duration": 213,
  "language": "en",
  "transcriptText": "We're no strangers to love You know the rules and so do I...",
  "srt": "1\n00:00:18,560 --> 00:00:20,560\nWe're no strangers to love\n\n2\n00:00:20,560 --> 00:00:22,560\nYou know the rules and so do I\n",
  "vtt": "WEBVTT\n\n00:00:18.560 --> 00:00:20.560\nWe're no strangers to love\n\n00:00:20.560 --> 00:00:22.560\nYou know the rules and so do I\n",
  "segments": [
    { "startTime": 18.56, "endTime": 20.56, "text": "We're no strangers to love" },
    { "startTime": 20.56, "endTime": 22.56, "text": "You know the rules and so do I" }
  ],
  "segmentCount": 54,
  "wordCount": 312,
  "scrapedAt": "2026-04-04T12:00:00.000Z",
  "error": null
}

No speech in the video:

{
  "videoUrl": "https://www.tiktok.com/@creator/video/1234567890",
  "platform": "tiktok",
  "videoId": "1234567890",
  "title": "My TikTok video",
  "transcriptText": null,
  "srt": null,
  "vtt": null,
  "segments": [],
  "segmentCount": 0,
  "wordCount": 0,
  "error": "No speech was detected in the video audio, so there is no transcript. No charge."
}

---

Tips for best results

• 🎯 Check transcriptSource — caption text comes from the platform; speech-to-text is AI output, so names and rare words can differ from what was said.
• 📋 Caption availability — Not all videos have captions. Auto-generated captions are available on most YouTube videos. TikTok and Instagram auto-captions are only available on a subset of videos (typically creators who have enabled them).
• 🗣️ Language selection is exact and YouTube-only — preferredLanguage selects an existing YouTube caption track; it does not translate or generate captions. If that exact language is unavailable, the result contains an honest error and no transcript extraction event is charged.
• 📦 Batch freely — mix platforms in one run. Speech-to-text takes about 30 seconds per short video, and the default run timeout is one hour.
• 🔄 Use maxRetries: 3 for large batches — transient failures are common with social media platforms. Set retries higher for unattended automation runs.
• 📅 Schedule for monitoring — use Apify's scheduler to run weekly transcript extraction on channels you monitor, and connect to Google Sheets for automatic updates.
• 🚀 Start small to test — run 5-10 URLs first to check the videos are public and contain speech.

---

Integrations

📊 Video Transcript Scraper → Google Sheets (content monitoring) Connect output via Apify webhooks to automatically append new transcripts to a Google Sheet. Use the built-in Google Sheets integration in Apify Console → Integrations.

🤖 Video Transcript Scraper → OpenAI / Claude (AI summarization) Feed transcriptText into an LLM prompt via Make.com or Zapier: "Summarize this video transcript in 3 bullet points and extract 5 key quotes." Build content briefs automatically from competitor videos.

📧 Video Transcript Scraper → Email alerts (keyword monitoring) Schedule weekly runs on a YouTube channel's latest videos. Use a Make.com filter on transcriptText to check for keyword mentions, then send a Slack or email alert when your brand or competitor is mentioned.

🗂️ Video Transcript Scraper → RAG pipeline (AI search) Export JSON output, chunk segments arrays by timestamp windows, embed with OpenAI, and index into Pinecone or Weaviate. Build a searchable knowledge base of video content queryable by natural language.

📝 Video Transcript Scraper → WordPress / CMS (SEO content) Pipe YouTube transcripts into a templating system that auto-generates blog post drafts. Each transcript becomes a lightly-edited article, turning a video library into an SEO content moat.

---

API usage

Run Video Transcript Scraper programmatically from any language or CI/CD pipeline.

Node.js:

import { ApifyClient } from 'apify-client';

const client = new ApifyClient({ token: 'YOUR_API_TOKEN' });

const run = await client.actor('automation-lab/video-transcript-scraper').call({
  videoUrls: [
    'https://www.youtube.com/watch?v=dQw4w9WgXcQ',
    'https://www.youtube.com/shorts/5-WE64bwEMI',
  ],
  maxRetries: 2,
});

const { items } = await client.dataset(run.defaultDatasetId).listItems();
console.log(items);

Python:

from apify_client import ApifyClient

client = ApifyClient(token="YOUR_API_TOKEN")

run = client.actor("automation-lab/video-transcript-scraper").call(run_input={
    "videoUrls": [
        "https://www.youtube.com/watch?v=dQw4w9WgXcQ",
        "https://www.tiktok.com/@nasa/video/7234567890",
    ],
    "maxRetries": 2,
})

items = client.dataset(run["defaultDatasetId"]).list_items().items
for item in items:
    print(item["platform"], item["wordCount"], item["transcriptText"][:100])

cURL:

curl -X POST \
  "https://api.apify.com/v2/acts/automation-lab~video-transcript-scraper/runs" \
  -H "Authorization: Bearer YOUR_API_TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "videoUrls": ["https://www.youtube.com/watch?v=dQw4w9WgXcQ"],
    "maxRetries": 2
  }'

---

Use with AI agents via MCP

Video Transcript Scraper is available as a tool for AI assistants that support the Model Context Protocol (MCP).

Add the Apify MCP server to your AI client — this gives you access to all Apify actors, including this one:

Setup for Claude Code

$claude mcp add --transport http apify "https://mcp.apify.com?tools=automation-lab/video-transcript-scraper"

Setup for Claude Desktop, Cursor, or VS Code

Add this to your MCP config file:

{
    "mcpServers": {
        "apify": {
            "url": "https://mcp.apify.com?tools=automation-lab/video-transcript-scraper"
        }
    }
}

Your AI assistant will use OAuth to authenticate with your Apify account on first use.

Example prompts

Once connected, try asking your AI assistant:

• "Use automation-lab/video-transcript-scraper to get the transcript of this YouTube video: https://www.youtube.com/watch?v=dQw4w9WgXcQ"
• "Extract transcripts from these 5 TikTok videos and summarize the main topics covered in each"
• "Get transcripts for all videos in this list and tell me which ones mention the keyword 'climate change'"

Learn more in the Apify MCP documentation.

---

Legality

This actor extracts publicly available caption data from public videos on social media platforms. It does not:

• Bypass any authentication walls
• Download video or audio content
• Access private or unlisted videos
• Store or redistribute video content

Caption data scraped is functionally equivalent to reading the closed captions a user can already see while watching a video.

Best practices:

• Only scrape public videos
• Respect platform terms of service for your use case
• Do not use transcript data to reproduce original creative works in full
• Review Apify's terms of service and the responsible scraping guidelines

This actor is intended for research, analysis, content repurposing, and automation workflows using publicly available data.

---

FAQ

Which platform has the highest transcript success rate? YouTube works for most videos with captions, including auto-generated ones. TikTok, Instagram and Facebook work for public videos that contain speech, because the audio is transcribed when there is no caption file. Music-only videos, private posts and photo posts return a free error row.

How much does it cost to transcribe 100 videos? Depends on the platform and length. On BRONZE, 100 YouTube videos cost $0.003 + (100 × $0.003) = $0.303. 100 Instagram Reels under a minute cost $0.003 + (100 × $0.0065) = $0.653. Videos longer than a minute cost one unit per started minute, and videos with no transcript are not charged.

How is this different from the official YouTube Transcript API? YouTube has no official transcript API. This actor uses the internal timedtext endpoint that the YouTube player uses — the same data that shows as closed captions when you watch a video. It works for any public video with captions, including auto-generated ones.

Why do some videos return no transcript? The video has no speech (music-only clips are common on TikTok and Reels), is private or deleted, or is longer than maxVideoDurationSeconds. The row's error field says which, and none of these are charged.

Can I extract transcripts from private or unlisted videos? No. This actor only works with publicly accessible videos. Private videos require authentication, which is outside the scope of this tool.

What languages are supported? For YouTube, any existing caption track can be requested with preferredLanguage (for example es, de, or pt-BR). Selection prefers an exact manual track and then an exact generated track. Leave the option empty for the existing English-first behavior, which prefers English and then falls back to the first available track. For TikTok, Instagram and Facebook, speech-to-text detects the spoken language by itself. preferredLanguage does not apply to those platforms, and it never translates or generates captions. If the requested YouTube language is unavailable, the result is an error row rather than a silent English fallback.

How fast does it run? YouTube transcripts take a few seconds each. Speech-to-text adds the download plus a few seconds for a short video, so a Reel usually finishes in well under a minute.

---

Privacy of diagnostics

Failed operations send sanitized diagnostic input, exceptions and Actor, build and run IDs to our private GlitchTip service for repair. Secret fields and URL queries are removed, and reports are retained for 30 days.

---

Other video and social media scrapers

Looking for related tools? Check out these other automation-lab actors:

• 🎵 TikTok Scraper — Extract TikTok videos, profiles and stats
• 📺 YouTube Scraper — Scrape YouTube video data, channel info, and comments
• 📸 Instagram Transcript Scraper — Instagram-only transcripts with source metadata
• 📝 YouTube Transcript Scraper — YouTube-only transcripts, including channels and playlists
• 🛒 Amazon Reviews Scraper — Extract product reviews from Amazon for sentiment analysis