Log inSubmit server
YouTube Research MCP logo

YouTube Research MCP

Researches YouTube transcripts with timestamped citations, plus optional search, comments, and channel analytics.

InstallRepositoryWebsite
Full README Compare alternatives Embed widget Claim this listing

Install YouTube Research MCP

Pick your client, then add it in one click or paste the config.

STDIONode.js
~/Library/Application Support/Claude/claude_desktop_config.json
claude_desktop_config.json
{
  "mcpServers": {
    "youtube-research-mcp": {
      "command": "npx",
      "args": [
        "-y",
        "@coyasong/youtube-mcp-server@latest"
      ]
    }
  }
}

Settings → Developer → Edit Config, merge this into mcpServers, then fully quit and reopen Claude Desktop.

Install not yet sandbox-tested— we haven't run this command through our automated sandbox yet. Not a red flag; we're working through the catalog.

Overview

YouTube Research MCP MCP server turns YouTube URLs or video IDs into structured transcript evidence with timestamps and links to the cited moments. It works without an API key for transcript research, filtering, key moments, and segmentation; adding a YouTube Data API v3 key enables search, comments, statistics, trends, and comparisons. Use it when an agent needs source-linked evidence from videos or limited channel and video intelligence rather than an uncited transcript dump. It supports local stdio use and a Streamable HTTP endpoint for remote deployments.

Use cases

  • Extract timestamped evidence from a public video
  • Compare claims across two to five videos
  • Search captions for a phrase and nearby context
  • Review channel and video performance statistics
  • Find trending videos by region and category

Key features

  • Citation-ready transcript evidence
  • Transcript filtering and segmentation
  • Cross-video evidence comparison
  • YouTube video and channel analytics
  • Optional comments and video search
  • Stdio and Streamable HTTP transports

Tools14

What YouTube Research MCP lets your agent do. Expand a tool for its parameters.

Self-reported~269 tokens of context

Security & health

What it can reach and how reliably it runs — from our own checks, not the maintainer’s claims.

ActiveLast checked 3d ago

Recent health check succeeded.

6/6 checks healthy over the last 40d

Supply-chain signal

No high-severity advisories surfaced by our automated scan.

Critical 0High 0Medium 0Low 0

Scanned 15d ago via OSV.dev · @coyasong/youtube-mcp-server@latest (npm)

Adoption & maintenance

Factual signals from GitHub, npm, and our automated checks — not a rating.

GitHub stars26Stargazers on the source repository.
npm downloads796Package downloads in the last 30 days.
Last commit3d agoMost recent push to the default branch.
Tools exposed14Callable tools this server registers over MCP.

How YouTube Research MCP works

What the YouTube Research MCP MCP server does

The YouTube Research MCP MCP server gives MCP clients tools for researching video captions and, optionally, querying YouTube Data API information. Transcript-focused operations accept standard YouTube URLs, short links, Shorts, embeds, live URLs, and raw 11-character video IDs. Results can include the video title, channel identity, caption duration, segment counts, matching text, timestamps, canonical links, and pagination details.

The service has two practical modes. Transcript-only mode needs no YouTube API key and covers caption research, transcript filtering, key-moment extraction, and segmentation. Full mode adds API-backed features such as video search, comments, video and channel statistics, trending videos, categories, cross-video comparisons, and channel video analysis.

How it works

An MCP client connects over local stdio or Streamable HTTP. The server validates tool inputs, retrieves caption tracks for transcript operations, and derives timestamp labels and links from caption segments. It keeps data in an in-memory cache. API-backed operations use YouTube Data API v3 when a key is available; otherwise, those operations return a key-required error.

research-video is intended for focused evidence gathering. A query can search within a transcript, while context-line, language, time-window, pagination, match-mode, and result-limit options control the returned material. research-videos applies one query to two through five videos and returns timestamp-linked results for each source. This approach can reduce the amount of transcript text an agent must process.

Setup and configuration

Node.js 20 or newer is required for the local installation. The package can run through npx over stdio, or it can be built from a clone and started as an HTTP service. The HTTP endpoint is /mcp, and the default port is 3000. A /health endpoint is available for checks.

YOUTUBE_API_KEY is optional. Set it to enable search, comments, and analytics. For a remote deployment, MCP_BEARER_TOKEN protects MCP requests, CORS_ORIGIN defines an exact comma-separated browser-origin allowlist, MAX_SESSIONS limits concurrent session memory, and SESSION_IDLE_TIMEOUT_MS controls idle-session cleanup. Browser origins are denied by default, and the health endpoint remains public.

Tools and capabilities

The YouTube Research MCP MCP server exposes tools for:

  • Searching and comparing transcript evidence across videos.
  • Retrieving complete or language-specific transcripts.
  • Filtering transcript text, extracting key moments, and dividing transcripts into segments.
  • Searching videos and retrieving comments when API mode is enabled.
  • Reading video and channel statistics, trends, and regional categories.
  • Comparing video statistics and analyzing performance trends for a channel’s videos.

get-trending-videos and get-video-categories accept regional context. Statistical and channel-analysis tools depend on the YouTube API key, while caption availability remains a prerequisite for transcript results.

Limitations and notes

A transcript cannot be returned when the requested video has no usable captions. Private, age-restricted, region-restricted, or caption-disabled videos may therefore fail transcript research. Transcript tools do not require an API key, but search, comments, metadata, statistics, trends, and comparison features do.

Remote full-mode deployments should be authenticated because unauthenticated access could consume the configured YouTube API quota. Citation links point to the corresponding time in the canonical YouTube video URL. For long videos, focused queries or time windows are preferable; pagination can retrieve additional evidence when needed.

Read the full README View source on GitHub

Reviews

No reviews yet — be the first to share how it worked for you.

Frequently Asked Questions about YouTube Research MCP

YouTube Research MCP is an MCP server that connects AI clients to YouTube caption tracks and, optionally, YouTube Data API v3. Its main tools research transcripts, compare evidence across videos, search videos, retrieve comments, extract key moments, and inspect video or channel statistics with timestamped citation links.

Install it locally with Node.js 20 or newer by running it through npx: `npx -y @coyasong/youtube-mcp-server@latest`. Configure the command for a stdio MCP client, or clone the repository, install dependencies, build it, and run the HTTP server.

No, transcript research does not require an API key. Set `YOUTUBE_API_KEY` to enable video search, comments, metadata, statistics, trends, and comparison features.

It returns transcript segments with human-readable timestamps and YouTube URLs that open at the cited moment. The links and timestamps are derived from caption segments, and focused queries can return only relevant evidence with nearby context.

Yes, it supports Cursor through the stdio setup using the package command and arguments. The README also identifies Cursor as a supported MCP client.

Related MCP servers

All in Social Media All alternatives
Add a badge to your READMEShow it’s listed — live SVG, Markdown or HTML.

AllMCPs Directory Badge

Full Badge Customizer

Showcase your server listing on GitHub or your project documentation. Embed this dynamic SVG badge to highlight official listing status and live engagement.

Badge Style:
Live Dynamic SVG PreviewYouTube Research MCP AllMCPs Directory Badge
Markdown (GitHub README)
[![AllMCPs](https://allmcps.com/api/badge/youtube-research-mcp?style=directory)](https://allmcps.com/mcp/youtube-research-mcp)
HTML Embed
<a href="https://allmcps.com/mcp/youtube-research-mcp"><img src="https://allmcps.com/api/badge/youtube-research-mcp?style=directory" alt="YouTube Research MCP on AllMCPs" /></a>