M
MCP Directory CURATED DIRECTORY
911 MCP servers indexed
中文
← Back to directory
Y

YouTube Data MCP Server

Community
High-efficiency YouTube MCP server for AI agents
How to install ↓ Compare GitHub source repository ↗
Category
Other #57 of 230
Stars
★ 32 Popular
Transport
stdio (local process)
Runtime
Node.js 20+
Credentials
Optional API key
License
MIT
Last commit
Tools
9
62FMRS · C

Users praise the server for its token optimization, significantly reducing context window usage. The caching mechanism effectively protects API quotas. Code quality is high with 97% test coverage, and it is actively maintained. Suitable for production, but note it's unofficial and potential quota costs for search operations.

Strongest · Setup experience 14/20 Weakest · Reliability 10/20

Reliability
10/20
Security and permissions
12/20
Maintenance
12/20
Documentation
14/20
Setup experience
14/20
Why each score
Reliability 10/20
README claims 97% coverage, CI badges, Zod validation and the official MCP starter, but these are untrusted README evidence; no actual test files or CI workflow contents were supplied for static verification. The tool list is coherent with detailed quota-cost disclosure, and degradation is partly documented (findConsistentOutlierChannels unavailable without MongoDB). Deducted for lack of verifiable execution evidence; capped by calibration rules.
Security and permissions 12/20
API key passed via env var marked isSecret, key restriction guidance given, no real tokens, malware or overreach found. Deducted because the MongoDB cache layer's data boundaries (remote cached content, access control) are not disclosed, and the multi-request outlier-analysis tool's network behavior is only loosely described. No red-line risk identified.
Maintenance 12/20
Active status, v1.1.13 aligned between manifest and npm, MIT license, claimed Dependabot patching and CI, 1 open issue. Deducted for absence of verifiable evidence on commit cadence, issue response time, release rhythm and a security-response channel; small project (26 stars) with incomplete governance visibility.
Documentation 14/20
Strong layered docs: install, API key setup, per-tool quota costs, Windows notes, local dev workflow, caching behavior, security considerations. Deducted for unverified self-reported token-savings numbers, an internal contradiction between the tool table (niche/minVideos/maxChannels) and the deep-dive section (query/channelAge/consistencyLevel) for findConsistentOutlierChannels, and missing troubleshooting content.
Setup experience 14/20
Zero-config mode, clear JSON config examples for Claude Desktop/Cursor, npx-based one-step launch, Windows PowerShell fallback, and .env.example template. Deducted because npx startup and the Node >=20 prerequisite are common fragility points, local dev requires an absolute working_directory, and static review cannot confirm the install path actually works; setup stays at or below the 15 cap per calibration rules.

Static review · not runListed 2026-09-08

Read the FMRS scoring method →

Fit and risk

What it can accessUses the networkConnects to a database

Best for

  • LLM agents that need to minimize token consumption.
  • Users who frequently access YouTube data and want to protect API quotas.
  • Developers seeking a high-quality, well-tested MCP server.

Not for

  • Users needing raw, unoptimized YouTube API responses with all nested fields.
  • Users requiring real-time or very fresh data (caching may introduce staleness).
  • Users without MongoDB who want to use the `findConsistentOutlierChannels` tool.

Required permissions

  • Network access to the YouTube Data API v3.
  • Optional: Access to API key from environment variables.
  • Optional: Access to MongoDB connection string for caching.

Risks and side effects

  • API key exposure: Stored in env vars, risk of accidental commit if misconfigured.
  • API quota exhaustion: Search tools still cost significant units (e.g., search 100 units) despite optimizations.
  • Data caching may return stale info; avoid enabling MongoDB if not needed.
  • Dependency on a third-party server (non-official), subject to changes as YouTube API evolves.

Setup

Before you start

Runtime：Node.js 20+

YOUTUBE_API_KEY optionalsecret YouTube Data API v3 key; enable the YouTube Data API v3 in Google Cloud Console and create an API key. Optional; unlocks search and channel analytics tools.
MDB_MCP_CONNECTION_STRING optionalsecret MongoDB connection string for caching and the findConsistentOutlierChannels tool; get a free one from MongoDB Atlas. Optional.
  1. Zero-config mode: Configure the MCP client with command npx -y @kirbah/mcp-youtube, no API key required for transcripts.
  2. Full features: Go to Google Cloud Console, enable YouTube Data API v3, and create an API key. Optionally, create a free MongoDB connection string for caching.
  3. Set environment variables YOUTUBE_API_KEY and MDB_MCP_CONNECTION_STRING in your client config.

Example config:

{
  "mcpServers": {
    "youtube": {
      "command": "npx",
      "args": ["-y", "@kirbah/mcp-youtube"],
      "env": {
        "YOUTUBE_API_KEY": "YOUR_KEY",
        "MDB_MCP_CONNECTION_STRING": "mongodb+srv://..."
      }
    }
  }
}
claude_desktop_config.json 
{
  "mcpServers": {
    "youtube": {
      "command": "npx",
      "args": ["-y", "@kirbah/mcp-youtube"]
    }
  }
}

Shown for Claude Desktop. Other clients may use a different file or key (VS Code uses "servers") — the configurator below converts it.

.vscode/mcp.json
{
  "servers": {
    "youtube": {
      "command": "npx",
      "args": [
        "-y",
        "@kirbah/mcp-youtube"
      ]
    }
  }
}

Goes in your project's .vscode/mcp.json (VS Code uses a "servers" key).

Terminal
claude mcp add youtube -- npx -y @kirbah/mcp-youtube

Run it in a terminal; replace any <…> placeholders with your own values first.

Generate and validate client config

Check that it works

Check that tools like getVideoDetails, searchVideos, and getTranscripts appear in the client's tool list, or ask the AI to read the transcript for a known video ID; a returned transcript confirms the connection works.

Troubleshooting

  1. Common issues:
  2. - If tools fail, check that `YOUTUBE_API_KEY` is correctly set and not expired.
  3. - Ensure network access to Google APIs.
  4. - If MongoDB caching enabled, verify the connection string is valid.
  5. - On quota errors, review usage in Google Cloud Console.

Things to try

Once connected, you can ask your AI assistant things like:

  • Read the transcript for youtube://transcript/dQw4w9WgXcQ
  • Search for videos about 'home espresso setup' and give me the top 5 results
  • Get the subscriber count, view count, and video count for channel UC_x5XG1OV2P6uZZ5FSM9Ttw
  • Show me the trending videos in the US right now

Tools 9

getVideoDetails read-only
Retrieves detailed, lean information for multiple YouTube videos including metadata, statistics, and engagement ratios.
searchVideos read-only
Searches for videos or channels based on a query string, returning concise results.
getTranscripts read-only
Fetches token-efficient transcripts (captions) for multiple videos, with options for full text or key segments.
getChannelStatistics read-only
Retrieves lean statistics for multiple channels (subscriber count, view count, video count, creation date).
getChannelTopVideos read-only
Retrieves a list of a channel's top-performing videos with lean details and engagement ratios.
getTrendingVideos read-only
Retrieves a list of trending videos for a region and optional category, with lean details and engagement ratios.
getVideoCategories read-only
Retrieves available YouTube video categories (ID and title) for a specific region.
getVideoComments read-only
Retrieves comments for a YouTube video, with sorting, limiting, and optional replies.
Show 1 more tools
findConsistentOutlierChannels read-only
Identifies channels that consistently outperform within a niche. Requires a MongoDB connection.

Use cases

Content creators analyze their own channel performance to optimize strategy.
Marketers research trending videos and popular channels.
Developers build AI applications that require video metadata or transcripts.
Researchers analyze comment sentiment and engagement metrics.

Supported clients

Claude Desktop
CursorPartial support

Listed from the project's documentation, not tested by this site.

Overview

The YouTube Data MCP Server (@kirbah/mcp-youtube) is a production-grade MCP server designed for AI agents, fetching structured data via the YouTube Data API v3. It optimizes token usage, saving up to 87% on context window, includes smart caching to protect API quotas, and boasts high test coverage. It provides tools like fetching video details, searching videos, retrieving transcripts, channel statistics, trending videos, and more. Supports zero-config mode (no API key needed for transcripts) and optional MongoDB caching for quota savings.

Similar servers

BioMCP 75 · B

One binary. One grammar. Evidence from the biomedical sources you already trust.

★ 659 · Tools 11 Compare with this →
Compare with other MCP servers

Source revision 595f3c22fbbc Data synced 2026-10-19 Read the FMRS scoring method