Created by Lexxer.

YouTube Transcript API & MCP Server

TranscriptPack is a low-cost YouTube transcript API for Apify Console, REST clients, automations, and MCP-connected tools. It handles public videos, Shorts, playlists, and channels without requiring a YouTube API key.

Get useful text instead of hundreds of tiny caption fragments. Choose the formats you need and receive predictable results for research, search, subtitles, RAG, and software workflows.

Why use TranscriptPack?

• Pay only for successful transcripts — unavailable, empty, and captionless videos are free.
• Ready-to-use output — normalized text, readable paragraphs, timestamps, and deterministic chunks with overlap.
• Bulk sources — process individual videos, Shorts, playlists, or complete channel pages.
• Reliable caption access — direct access is tried first, with residential fallback when YouTube blocks data-center traffic.
• Multiple export formats — plain text, Markdown, SRT, WebVTT, segments, paragraphs, and chunks.
• Built for automation — stable fields, clear error codes, content hashes, and an OpenAPI-documented endpoint.
• Native MCP tool — connect any Streamable HTTP MCP client and call get_youtube_transcripts directly.

Pricing

$0.0005 per successfully delivered transcript — $0.50 per 1,000.

There is no start fee and failed transcripts are not charged. Apify platform usage is billed separately by Apify and may include compute, storage, transfer, and residential proxy traffic when fallback is needed.

Quick start

Send a POST request to /v1/transcripts with a video URL:

{
  "startUrls": [
    "https://www.youtube.com/watch?v=neabI31ofMc"
  ],
  "languages": ["en"],
  "formats": ["text", "chunks", "markdown"],
  "maxVideos": 1,
  "proxyMode": "residential-fallback"
}

Use the Endpoints tab to try the API directly or copy the generated request for your application.

Connect with MCP

TranscriptPack exposes a native Streamable HTTP MCP server at:

https://lexxer--transcriptpack-youtube-transcripts.apify.actor/mcp

Authenticate with your Apify API token as a Bearer token. MCP clients discover one focused tool: get_youtube_transcripts.

You can also find TranscriptPack through Apify's hosted MCP server and call lexxer/transcriptpack-youtube-transcripts:get_youtube_transcripts.

What you receive

Each successful result can include:

• video ID, URL, title, author, and thumbnail;
• selected caption language and translation details;
• normalized transcript text;
• timestamped caption segments;
• readable timestamped paragraphs;
• deterministic chunks with configurable size and overlap;
• Markdown, SRT, and WebVTT exports;
• word, character, and snippet counts;
• a SHA-256 content hash for caching and change detection;
• proxy usage information.

Only the formats you request are returned, keeping responses compact.

Process playlists and channels

Provide several source URLs in one request or use a playlist or channel page. TranscriptPack resolves the videos, removes duplicates, applies the global maxVideos limit, and returns one structured result per video.

{
  "startUrls": [
    "https://www.youtube.com/playlist?list=PLBkiWvdL-IJa8NvXNuJpKiUEwQ5UHHw8t"
  ],
  "languages": ["en", "de"],
  "formats": ["text", "chunks"],
  "maxVideos": 25,
  "concurrency": 3
}

Common uses

• Build a searchable YouTube knowledge base.
• Prepare videos for RAG and embedding pipelines.
• Summarize long videos, playlists, or channel archives.
• Export subtitles as SRT or WebVTT.
• Monitor transcript changes with content hashes.
• Feed research and content-analysis workflows.
• Collect multilingual captions using YouTube-provided translations.

Smart access modes

• residential-fallback — recommended; tries direct access first and falls back only when needed.
• direct — lowest platform usage, but more likely to be blocked by YouTube.
• residential — uses residential access immediately for maximum consistency.

An optional two-letter country code can target residential traffic to a specific country.

Supported content

• Public YouTube videos and Shorts
• Public playlists and channel pages
• Manual and auto-generated captions
• YouTube-provided caption translations
• Up to 500 resolved videos per request

TranscriptPack uses existing YouTube captions. It does not transcribe audio, access private videos, bypass account restrictions, or guarantee captions for unavailable, age-restricted, or captionless content.

Responsible use

Use TranscriptPack only for content you may lawfully process. Caption accuracy varies, and users are responsible for their inputs and downstream use.