YouTube Vision MCP

Example Use Cases

• 1

  Answer user questions about the content of a YouTube video using vision-enabled model analysis.

• 2

  Generate concise or extended summaries of video content for indexing, previews, or notes.

• 3

  Extract and present key moments (timestamps and short descriptions) from videos for navigation or highlight generation.

Description

This MCP server connects to the Google Gemini Vision API to analyze YouTube video content and expose tools for Q&A, summaries, and key-moment extraction. It provides dedicated tools (ask_about_youtube_video, summarize_youtube_video, extract_key_moments, list_supported_models) and is configurable via environment variables (GEMINI_API_KEY, GEMINI_MODEL_NAME). The server communicates over stdio so it can be integrated into MCP clients (e.g., Claude, VSCode) and is distributed as an open-source npm package. It is useful for automating video understanding, indexing, and conversational workflows around video content.