MCP server that retrieves captions/subtitles from YouTube videos via a simple Model Context Protocol interface.
https://github.com/kimtaeyoon83/mcp-server-youtube-transcriptTired of manually grabbing YouTube subtitles for your AI projects? This MCP server eliminates that workflow friction entirely by giving Claude (and other MCP clients) direct access to YouTube video transcripts.
You know the drill: find a video with valuable content, manually enable captions, copy chunks of text, paste into your AI chat, repeat. It's tedious, error-prone, and breaks your flow when you're trying to analyze educational content, extract key insights, or build applications that need video content understanding.
One simple tool: get_transcript. Pass it any YouTube URL format (full URLs, youtu.be links, or raw video IDs) and get back clean, timestamped transcript data. Works across languages without needing API keys or quota management - it uses direct caption scraping.
# In Claude Desktop, it's literally this simple:
# "Extract transcript from https://youtu.be/ODaHJzOyVCQ"
Content Research: Analyzing hour-long technical talks or educational videos becomes instant. Instead of watching and taking notes, get the full transcript and have Claude extract key concepts, action items, or technical details.
Educational Workflows: Building study materials from lecture videos? Get transcripts in seconds, then have AI create summaries, flashcards, or concept maps from the content.
Application Development: Need video content for your AI app? This server provides clean, structured transcript data you can process programmatically without dealing with YouTube's API complexity.
Multi-language Content: Supports language-specific transcript retrieval (lang parameter), so you can work with content in Korean, Spanish, French, or any language with available captions.
npx -y @smithery/cli install @kimtaeyoon83/mcp-server-youtube-transcript --client claude
Or add directly to your Claude Desktop config:
{
"mcpServers": {
"youtube-transcript": {
"command": "npx",
"args": ["-y", "@kimtaeyoon83/mcp-server-youtube-transcript"]
}
}
}
Instead of context-switching between YouTube and your AI tools, video transcripts become just another data source in your workflow. Research educational content, analyze technical presentations, or extract insights from interviews without the manual copy-paste dance.
The 241 GitHub stars suggest developers are already finding this useful for eliminating a common productivity bottleneck. When you need video content accessible to AI, this server makes it seamless.
Ready to stop manually copying subtitles? Install it and start pulling YouTube transcripts directly into your AI workflows.