YouTube MCP. Audit channel growth or analyze video performance instantly.

Yo, forget manually clicking through YouTube Studio. This server lets your AI client run deep analytics on video data and channel performance straight from a chat command. You just talk to it, and it handles all the API junk.

When you need to find content, you use search_videos. Just toss in a keyword or phrase, and it hits YouTube's catalog, spitting out a list of titles and descriptions for videos that match your query. It gives you enough metadata—the title, the description, and the ID—so your agent knows exactly what it's looking at before it pulls any stats.

If you need to check out a specific video in detail, get_video does the heavy lifting. Give it just the video ID, and it pulls every piece of technical metadata available. That means views, likes, total watch time, the upload date, and the full description—everything's right there in one dump.

It’s way faster than digging through multiple tabs.

Want to audit a channel? You fire up get_channel. This tool gives you the whole picture on any specific YouTube channel ID. It pulls comprehensive stats, including the total subscriber count and how many videos the channel's put out. Plus, it grabs general branding information so you know what you're dealing with right off the bat.

For user sentiment, list_comments is your go-to. Feed it a video ID, and it fetches the most relevant or recent comment threads attached to that specific piece of content. Your AI client can then run through those comments, analyzing user feedback for tone and topic—you don't gotta read hundreds of replies yourself.

Here’s how you use this stack together. You might start by running search_videos on a niche topic, getting a list of candidate video IDs. Then, if one title looks promising, you pass that ID to get_video to pull the raw performance numbers—the views count and likes total, for example. If those stats look good, maybe you run list_comments on it to see what people are actually saying about the content.

And if you're building out a whole competitive landscape report, you can use get_channel to benchmark the entire channel against its peers by checking their subscriber count and video output volume. You also get unique identifiers for all these pieces of media—the specific IDs needed to structure your monitoring workflow.

This setup means that instead of spending time navigating YouTube's dashboard, you just talk to your agent. It handles getting the data from the search results, pulling deep stats on any single video, auditing entire channel health metrics, and analyzing user feedback from comment threads—all without you lifting a finger.

You get clean, structured data outputs for every tool call, letting your AI client chew through raw intelligence instantly.