YouTube Transcript Server

What is YouTube Transcript Server?

YouTube Transcript Server is a Model Context Protocol (MCP) server that retrieves transcripts from YouTube videos. It provides direct access to video captions and subtitles through a simple tool interface, supporting various video URL formats including YouTube Shorts.

How to use YouTube Transcript Server?

Install via Smithery (npx -y @smithery/cli install @kimtaeyoon83/mcp-server-youtube-transcript --client claude) or add the server configuration to your Claude Desktop settings. Invoke the get_transcript tool with a YouTube URL or video ID; optional parameters include lang, include_timestamps, and strip_ads.

Key features of YouTube Transcript Server

• Support for multiple video URL formats (including YouTube Shorts)
• Language-specific transcript retrieval with automatic fallback
• Optional timestamps for referencing specific moments
• Built-in ad/sponsorship filtering (enabled by default)
• Zero external dependencies for transcript fetching
• Detailed metadata in responses

Use cases of YouTube Transcript Server

• Extracting subtitles from YouTube videos for content analysis
• Obtaining time‑stamped quotes for research or citation
• Getting clean transcripts without sponsor or ad interruptions
• Enabling accessibility by converting video speech to text
• Feeding transcripts into LLMs for summarization or Q&A

FAQ from YouTube Transcript Server

What video URL formats are supported?

You can use full YouTube URLs, YouTube Shorts URLs, or just a video ID.

How does language fallback work?

If the requested language code (e.g., 'ko') is not available, the server automatically falls back to any other available transcript language.

How can I include timestamps in the transcript?

Set include_timestamps to true in the get_transcript call. Timestamps are formatted like [0:05] text.

Can I disable the ad/sponsorship filtering?

Yes. Set the strip_ads parameter to false to receive the raw, unfiltered transcript.