Youtube Transcripts MCP server

🎬 YouTube Transcripts API: Captions and Subtitles in Clean JSON

The efficient, reliable, and developer-friendly way to use the YouTube Transcripts API.

Actor page: apify.com/johnvc/YoutubeTranscripts Input schema: apify.com/johnvc/YoutubeTranscripts/input-schema

The YouTube Transcripts API extracts the transcript for one or more YouTube videos and returns clean, structured JSON. Each video comes back with a timestamped caption list (text, start, and duration), a plain-text version of the full transcript, and language metadata. It works with standard videos, Shorts, youtu.be short links, embed URLs, and mobile URLs, processes multiple URLs in parallel, and is billed per video.

Video Walkthrough

Quick Start

Prerequisites

• Python 3.11 or higher
• An Apify account and API key (get a free key here)

1. Clone the repository

git clone https://github.com/johnisanerd/Apify-Youtube-Transcripts-API.git
   cd Apify-Youtube-Transcripts-API

1. Install dependencies with UV

# Install UV if you do not have it:
   curl -LsSf https://astral.sh/uv/install.sh | sh

   # Install project dependencies:
   uv sync

1. Configure your API key

cp .env.example .env
   # Edit .env and add your Apify API key
   # Get your free API key at: https://apify.com?fpr=9n7kx3

1. Run the example

uv run python youtube-transcripts-api.py

Alternative: set the API key directly

export APIFY_API_TOKEN="your_api_key_here"
uv run python youtube-transcripts-api.py

Why Use This YouTube Transcripts API?

Two transcript formats. Each video returns a timestamped caption list and a plain-text version, so you can build interactive players or feed clean text straight into search and analysis.

Many URL formats. Standard videos, Shorts, youtu.be short links, embed URLs, and mobile URLs all work.

Batch in parallel. Pass an array of URLs and each is processed in parallel; invalid URLs are recorded as errors in the dataset without stopping the run.

Language metadata. Every record reports the transcript language, language code, and whether the captions were auto-generated.

Predictable, pay-per-use pricing. Billing is per video, not per second, with no subscription, which keeps both single and batch runs inexpensive.

Easy to automate. Call it from Python in a few lines, or load it as an MCP tool so assistants like Claude and Cursor can pull transcripts for you on demand.

Features

Core Capabilities

• Single or batch input: one URL as a string, or many as an array
• Broad URL support: standard videos, Shorts, youtu.be, embed, and mobile URLs
• Timestamped captions with text, start, and duration per segment
• Plain-text transcript as a single string
• Per-video error handling: invalid URLs are logged without failing the run

Data Quality

• Two formats per video: timestamped and plain text
• Language metadata: language name, language code, and auto-generated flag
• Video duration in seconds
• Success flag on every record for easy filtering
• Consistent JSON shape across every video

Usage Examples

Single video

{
  "youtube_url": "https://www.youtube.com/watch?v=aircAruvnKk"
}

Multiple videos (batch)

{
  "youtube_url": [
    "https://www.youtube.com/watch?v=aircAruvnKk",
    "https://youtu.be/9bZkp7q19f0",
    "https://www.youtube.com/shorts/abcdEFGhijk"
  ]
}

Input Parameters

Shortened. The full README is on GitHub.