YouTube MCP Server
Connect to the YouTube MCP server to search videos, pull video details and subtitles, read comments, and analyze channels and playlists using AI agents on Gumloop, Claude, or Cursor.
Try YouTube now
Type what you want done. Sign in and run it live with an AI agent.
|
YouTube
Installation
Get StartedCreate a Gumloop Account
To use this MCP, you need a Gumloop account. If you don't have one yet, you can create one for free.
Copy Your Server URL
Copy your MCP server URL and add it to your client. You'll be prompted to authorize on first use.
Tools (6)
Search Videos
3 per itemSearch YouTube videos by search terms. Returns video details including titles, views, likes, duration, channel info, and metadata.
Get Video Details
4 per itemGet detailed information about specific YouTube videos by URL or video ID. Returns comprehensive video metadata, statistics, and content details.
Get Channel Videos
3 per itemGet videos from a specific YouTube channel. Returns channel videos with full metadata and statistics. Supports fetching regular videos, Shorts, and Streams separately using max_results, max_shorts, and max_streams. Set max_results=0 to fetch only Shorts/Streams.
Get Playlist Videos
3 per itemGet videos from a YouTube playlist. Returns all videos in the playlist with metadata and statistics.
Get Channel Details
5 per itemGet detailed information about a YouTube channel including subscriber count, total videos, description, and metadata. Note: the response may include an 'error: NO_RESULTS' field which indicates no videos were collected (this is expected — channel metadata is still returned successfully). To collect videos, use get_channel_videos or search_videos instead.
Get Video Comments
5 per itemGet comments from YouTube videos. Returns comment text, author information, vote counts, reply counts, and metadata.
What is YouTube MCP?
The YouTube MCP server gives AI agents access to public YouTube data. That means agents can search videos, pull detailed video metadata and statistics, read subtitles and comments, and explore channels, playlists, Shorts, and Streams. It turns YouTube into a research source an AI agent can query directly, without you opening a browser and clicking through results.
If you spend time researching what’s working on YouTube, copying view and like counts into sheets, reading through hundreds of comments to gauge sentiment, or pulling transcripts to summarize videos, an AI agent can take that work off your plate. Describe what you need, and your AI agent will handle the YouTube research for you.
MCP stands for Model Context Protocol. It’s an open standard that gives AI agents a way to connect to external tools and services. There’s no YouTube account to link and no API key to manage, because the connection is built into Gumloop. After that, you can research videos, channels, and comments just by chatting with your AI agent.
Related MCP servers
TikTok MCP Server
Interact with TikTok to get videos, user profiles, hashtag content, and more using Apify's TikTok scraper
Instagram MCP Server
Interact with Instagram to get posts, comments, profiles, and more using Apify's Instagram scraper
X MCP Server
Search and retrieve tweets from X (Formerly Twitter) with comprehensive timeline, bookmark, and search capabilities
Reddit MCP Server MCP Server
Interact with Reddit using the Reddit API
Google Sheets MCP Server
Read, write and update your Google Sheets
Slack MCP Server
Interact with Slack channels and messages
What you can do with YouTube MCP on Gumloop
Search YouTube and discover videos
Search by keyword and get back matching videos with titles, view and like counts, duration, channel info, and other metadata. An AI agent can find the most relevant or highest-performing videos on any topic in seconds.
Pull detailed video metadata and statistics
Get comprehensive details for any video by URL or ID, including statistics and content details. Your AI agent can compare videos, track performance, or build a structured dataset from a list of links.
Get video subtitles and captions
Pull a video’s subtitles inline when you fetch search results or video details, for videos that have captions available. An AI agent can read, summarize, translate, or extract key points from a video without anyone watching it end to end.
Read and analyze comments
Pull comments from a video, including the comment text, author, vote counts, and reply counts. An agent can gauge sentiment, surface common questions, or roll up audience feedback into a clean summary.
Explore channels, Shorts, and Streams
Get channel details like subscriber count, total videos, and description, and pull a channel’s videos with full stats. You can fetch regular videos, Shorts, and Streams separately to analyze each format on its own.
Pull entire playlists
Get every video in a playlist with metadata and statistics, so an AI agent can analyze a series, a course, or a curated collection as a single dataset.
How to connect the Gumloop YouTube MCP Server
- 1
Create a free Gumloop account
Sign up at gumloop.com. No credit card required.
- 2
Add the YouTube MCP server
Copy your MCP server URL from Gumloop and add it to your preferred client (Claude, Cursor, or Gumloop workflows). You'll authorize on first use.
- 3
Start using YouTube in your AI workflows
That's it. Your AI agent can now search videos, pull details and subtitles, read comments, and analyze channels and playlists. Use it inside a Gumloop automation, in Claude Desktop, or in Cursor.
YouTube MCP use cases
Content and competitor research for marketers
Search a topic, pull the top videos with their view and like counts, and gather channel context, then have an AI agent summarize what formats and angles are performing. Marketing teams see what’s working on YouTube without manually scrubbing search results.
Audience sentiment analysis for social teams
Pull comments from your own or a competitor’s videos, run them through an AI model for sentiment and themes, and compile the highlights into a Slack message or a Google Sheet. Social teams catch what audiences love and complain about without reading every comment.
Video summarization and repurposing for creators
Fetch a video’s subtitles and details, then have an AI agent turn the transcript into a blog post, a thread, show notes, or a list of clip-worthy moments. Creators can repurpose one video into a week of content.
Channel monitoring for brand and PR
On a schedule, pull a channel’s latest videos, Shorts, and Streams with stats, and have an AI agent flag new uploads or spikes in views and post an alert in Slack. Brand teams can stay on top of what creators and competitors are publishing.
Trend and keyword research for SEO
Search a set of keywords, compare view counts, likes, and durations across the results, and have an agent write a ranked trend report to Google Sheets. SEO and content teams can ground their strategy in real YouTube demand signals.
Why use Gumloop for YouTube MCP
No YouTube account or API key to manage
Most YouTube integrations you’ll find on GitHub make you set up a Google Cloud project, enable the YouTube Data API, manage OAuth or an API key, and handle quotas yourself. With Gumloop the connection is built in, so you don’t provide a YouTube account, a Google login, or any key. Just add the server and start querying.
Works with multiple MCP clients
Use the YouTube MCP server endpoint in Claude Desktop, Cursor, or directly inside Gumloop agents. Same server URL, works with any MCP client.
Chain YouTube with 100+ other integrations
Combine YouTube with Google Sheets, Slack, Notion, and other MCP tools in a single AI agent. An agent can pull a video’s subtitles, summarize them with an LLM, and write the recap to a doc or post it to a channel.
Enterprise-grade and scalable
Built for teams, with role-based permissions and dedicated support for Pro users. For details on Gumloop’s security practices, see trust.gumloop.com.
Pricing includes a free plan
You can test the YouTube MCP integration on Gumloop’s free tier before committing. Paid plans start at $37/month.
Frequently asked questions
Ship YouTube agents in minutes
Connect any AI agent to 100+ MCP servers, fully hosted, zero setup.