How to integrate Youtube MCP with Vercel AI SDK v6

Trusted by
AWS
Glean
Zoom
Airtable
GET STARTED FOR FREEGET A DEMO

30 min · no commitment · see it on your stack

Youtube logo
Vercel AI SDK logo
divider

Introduction

This guide walks you through connecting Youtube to Vercel AI SDK v6 using the Composio tool router. By the end, you'll have a working Youtube agent that can list your most recent uploaded videos, get subscriber count for your channel, search youtube for trending tutorials through natural language commands.

This guide will help you understand how to give your Vercel AI SDK agent real control over a Youtube account through Composio's Youtube MCP server.

Before we dive in, let's take a quick look at the key ideas and tools involved.

Also integrate Youtube with

TL;DR

Here's what you'll learn:
  • How to set up and configure a Vercel AI SDK agent with Youtube integration
  • Using Composio's Tool Router to dynamically load and access Youtube tools
  • Creating an MCP client connection using HTTP transport
  • Building an interactive CLI chat interface with conversation history management
  • Handling tool calls and results within the Vercel AI SDK framework

What is Vercel AI SDK?

The Vercel AI SDK is a TypeScript library for building AI-powered applications. It provides tools for creating agents that can use external services and maintain conversation state.

Key features include:

  • streamText: Core function for streaming responses with real-time tool support
  • MCP Client: Built-in support for Model Context Protocol via @ai-sdk/mcp
  • Step Counting: Control multi-step tool execution with stopWhen: stepCountIs()
  • OpenAI Provider: Native integration with OpenAI models

What is the Youtube MCP server, and what's possible with it?

The Youtube MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Youtube account. It provides structured and secure access to your channel data, so your agent can perform actions like searching videos, managing playlists, retrieving channel insights, and handling subscriptions on your behalf.

  • Channel activity monitoring: Let your agent fetch and summarize recent channel activities, including uploads, likes, playlist additions, and more, to keep you up to date at a glance.
  • Automated video and playlist management: Easily list videos from any channel, retrieve your own playlists, and organize your content—all through AI-driven commands.
  • Channel analytics and statistics: Ask your agent to pull detailed channel metrics such as subscriber counts, total views, or video counts for quick reporting and insights.
  • Subscription management: Have your agent list your current subscriptions or even subscribe you to new channels based on your interests or instructions.
  • Search and caption handling: Empower your agent to search YouTube for videos, channels, or playlists, as well as retrieve and download caption tracks for accessible viewing and content repurposing.

Supported Tools & Triggers

Tools
Triggers
Add Video to PlaylistTool to add a video to a playlist by inserting a playlist item.
Insert Channel SectionTool to create a new channel section for the authenticated user's YouTube channel.
Insert Comment ReplyTool to create a reply to an existing YouTube comment.
Create PlaylistTool to create a new YouTube playlist on the authenticated user's channel.
Delete Channel SectionTool to delete a YouTube channel section.
Delete CommentTool to delete a YouTube comment owned by the authenticated user or channel.
Delete PlaylistTool to delete a YouTube playlist owned by the authenticated user/channel.
Delete Playlist ItemTool to delete a playlist item (remove a video from a playlist).
Delete VideoTool to delete a YouTube video owned by the authenticated user/channel.
Get Channel ActivitiesGets recent activities from a YouTube channel including video uploads, playlist additions, likes, and other channel events.
Get channel ID by handleRetrieves the YouTube Channel ID for a specific YouTube channel handle.
Get Channel StatisticsGets detailed statistics for YouTube channels including subscriber counts, view counts, and video counts.
Video Details BatchRetrieves multiple YouTube video resource parts in a single batch call.
Get Video RatingRetrieves the ratings that the authorized user gave to a list of specified videos.
List captionsRetrieves a list of caption tracks for a YouTube video.
List Channel SectionsTool to retrieve channel sections from YouTube.
List channel videosLists videos from a specified YouTube channel.
List CommentsList individual comments from YouTube videos.
List Comment ThreadsTool to retrieve comment threads from YouTube videos or channels matching API request parameters.
List I18n LanguagesReturns a list of application languages that the YouTube website supports.
List I18n RegionsTool to retrieve a list of content regions that the YouTube website supports.
List Live Chat MessagesTool to list live chat messages for a specific chat.
List Playlist ImagesTool to retrieve playlist images associated with a specific playlist.
List Playlist ItemsTool to list videos in a playlist, with pagination support.
List Super Chat EventsLists Super Chat events for a channel, showing supporter purchases during live streams.
List user playlistsRetrieves playlists owned by the authenticated user, implicitly using mine=True.
List user subscriptionsRetrieves the authenticated user's YouTube channel subscriptions, allowing specification of response parts and pagination.
List Video Abuse Report ReasonsTool to retrieve a list of abuse report reasons that can be used to report abusive videos on YouTube.
List Video CategoriesTool to list YouTube video categories that can be associated with videos.
Download YouTube caption trackDownloads a specific YouTube caption track, which must be owned by the authenticated user, and returns its content as text.
Multipart upload videoUploads a video to YouTube using multipart upload in a single request.
Post Comment on VideoTool to post a new top-level comment on a YouTube video.
Rate VideoTool to add a like or dislike rating to a YouTube video, or remove an existing rating.
Report Video for AbuseTool to report a YouTube video for containing abusive content.
Search YouTubeSearches YouTube for videos, channels, or playlists using a query term, returning the raw API response.
Set Comment Moderation StatusTool to set the moderation status of one or more YouTube comments.
Subscribe to channelSubscribes the authenticated user to a specified YouTube channel, identified by its unique `channelId` which must be valid and existing.
Unsubscribe from channelTool to unsubscribe the authenticated user from a YouTube channel by deleting a subscription.
Update caption trackUpdates a YouTube caption track's metadata such as name, language, or draft status.
Update channelUpdates a channel's metadata including branding settings and localizations.
Update Channel SectionTool to update an existing YouTube channel section by ID.
Update CommentTool to modify the text of an existing YouTube comment.
Update PlaylistTool to modify an existing YouTube playlist's metadata (title, description, privacy status).
Update Playlist ItemTool to modify a playlist item's properties such as position or note.
Update thumbnailSets the custom thumbnail for a YouTube video using an image from a URL.
Update videoUpdates metadata for a YouTube video identified by videoId, which must exist; an empty list for tags removes all existing tags.
Upload videoUploads a video from a local file path to a YouTube channel; the video file must be in a YouTube-supported format.

What is the Composio tool router, and how does it fit here?

What is Composio SDK?

Composio's Composio SDK helps agents find the right tools for a task at runtime. You can plug in multiple toolkits (like Gmail, HubSpot, and GitHub), and the agent will identify the relevant app and action to complete multi-step workflows. This can reduce token usage and improve the reliability of tool calls. Read more here: Getting started with Composio SDK

The tool router generates a secure MCP URL that your agents can access to perform actions.

How the Composio SDK works

The Composio SDK follows a three-phase workflow:

  1. Discovery: Searches for tools matching your task and returns relevant toolkits with their details.
  2. Authentication: Checks for active connections. If missing, creates an auth config and returns a connection URL via Auth Link.
  3. Execution: Executes the action using the authenticated connection.

Step-by-step Guide

Prerequisites

Before you begin, make sure you have:
  • Node.js and npm installed
  • A Composio account with API key
  • An OpenAI API key

Getting API Keys for OpenAI and Composio

OpenAI API Key
  • Go to the OpenAI dashboard and create an API key. You'll need credits to use the models, or you can connect to another model provider.
  • Keep the API key safe.
Composio API Key
  • Log in to the Composio dashboard.
  • Navigate to your API settings and generate a new API key.
  • Store this key securely as you'll need it for authentication.

Install required dependencies

bash
npm install @ai-sdk/openai @ai-sdk/mcp @composio/core ai dotenv

First, install the necessary packages for your project.

What you're installing:

  • @ai-sdk/openai: Vercel AI SDK's OpenAI provider
  • @ai-sdk/mcp: MCP client for Vercel AI SDK
  • @composio/core: Composio SDK for tool integration
  • ai: Core Vercel AI SDK
  • dotenv: Environment variable management

Set up environment variables

bash
OPENAI_API_KEY=your_openai_api_key_here
COMPOSIO_API_KEY=your_composio_api_key_here
COMPOSIO_USER_ID=your_user_id_here

Create a .env file in your project root.

What's needed:

  • OPENAI_API_KEY: Your OpenAI API key for GPT model access
  • COMPOSIO_API_KEY: Your Composio API key for tool access
  • COMPOSIO_USER_ID: A unique identifier for the user session

Import required modules and validate environment

typescript
import "dotenv/config";
import { openai } from "@ai-sdk/openai";
import { Composio } from "@composio/core";
import * as readline from "readline";
import { streamText, type ModelMessage, stepCountIs } from "ai";
import { createMCPClient } from "@ai-sdk/mcp";

const composioAPIKey = process.env.COMPOSIO_API_KEY;
const composioUserID = process.env.COMPOSIO_USER_ID;

if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set");
if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set");

const composio = new Composio({
  apiKey: composioAPIKey,
});
What's happening:
  • We're importing all necessary libraries including Vercel AI SDK's OpenAI provider and Composio
  • The dotenv/config import automatically loads environment variables
  • The MCP client import enables connection to Composio's tool server

Create Tool Router session and initialize MCP client

typescript
async function main() {
  // Create a tool router session for the user
  const session = await composio.create(composioUserID!, {
    toolkits: ["youtube"],
  });

  const mcpUrl = session.mcp.url;
What's happening:
  • We're creating a Tool Router session that gives your agent access to Youtube tools
  • The create method takes the user ID and specifies which toolkits should be available
  • The returned mcp object contains the URL and authentication headers needed to connect to the MCP server
  • This session provides access to all Youtube-related tools through the MCP protocol

Connect to MCP server and retrieve tools

typescript
const mcpClient = await createMCPClient({
  transport: {
    type: "http",
    url: mcpUrl,
    headers: session.mcp.headers, // Authentication headers for the Composio MCP server
  },
});

const tools = await mcpClient.tools();
What's happening:
  • We're creating an MCP client that connects to our Composio Tool Router session via HTTP
  • The mcp.url provides the endpoint, and mcp.headers contains authentication credentials
  • The type: "http" is important - Composio requires HTTP transport
  • tools() retrieves all available Youtube tools that the agent can use

Initialize conversation and CLI interface

typescript
let messages: ModelMessage[] = [];

console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
console.log(
  "Ask any questions related to youtube, like summarize my last 5 emails, send an email, etc... :)))\n",
);

const rl = readline.createInterface({
  input: process.stdin,
  output: process.stdout,
  prompt: "> ",
});

rl.prompt();
What's happening:
  • We initialize an empty messages array to maintain conversation history
  • A readline interface is created to accept user input from the command line
  • Instructions are displayed to guide the user on how to interact with the agent

Handle user input and stream responses with real-time tool feedback

typescript
rl.on("line", async (userInput: string) => {
  const trimmedInput = userInput.trim();

  if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) {
    console.log("\nGoodbye!");
    rl.close();
    process.exit(0);
  }

  if (!trimmedInput) {
    rl.prompt();
    return;
  }

  messages.push({ role: "user", content: trimmedInput });
  console.log("\nAgent is thinking...\n");

  try {
    const stream = streamText({
      model: openai("gpt-5"),
      messages,
      tools,
      toolChoice: "auto",
      stopWhen: stepCountIs(10),
      onStepFinish: (step) => {
        for (const toolCall of step.toolCalls) {
          console.log(`[Using tool: ${toolCall.toolName}]`);
          }
          if (step.toolCalls.length > 0) {
            console.log(""); // Add space after tool calls
          }
        },
      });

      for await (const chunk of stream.textStream) {
        process.stdout.write(chunk);
      }

      console.log("\n\n---\n");

      // Get final result for message history
      const response = await stream.response;
      if (response?.messages?.length) {
        messages.push(...response.messages);
      }
    } catch (error) {
      console.error("\nAn error occurred while talking to the agent:");
      console.error(error);
      console.log(
        "\nYou can try again or restart the app if it keeps happening.\n",
      );
    } finally {
      rl.prompt();
    }
  });

  rl.on("close", async () => {
    await mcpClient.close();
    console.log("\n👋 Session ended.");
    process.exit(0);
  });
}

main().catch((err) => {
  console.error("Fatal error:", err);
  process.exit(1);
});
What's happening:
  • We use streamText instead of generateText to stream responses in real-time
  • toolChoice: "auto" allows the model to decide when to use Youtube tools
  • stopWhen: stepCountIs(10) allows up to 10 steps for complex multi-tool operations
  • onStepFinish callback displays which tools are being used in real-time
  • We iterate through the text stream to create a typewriter effect as the agent responds
  • The complete response is added to conversation history to maintain context
  • Errors are caught and displayed with helpful retry suggestions

Complete Code

Here's the complete code to get you started with Youtube and Vercel AI SDK:

typescript
import "dotenv/config";
import { openai } from "@ai-sdk/openai";
import { Composio } from "@composio/core";
import * as readline from "readline";
import { streamText, type ModelMessage, stepCountIs } from "ai";
import { createMCPClient } from "@ai-sdk/mcp";

const composioAPIKey = process.env.COMPOSIO_API_KEY;
const composioUserID = process.env.COMPOSIO_USER_ID;

if (!process.env.OPENAI_API_KEY) throw new Error("OPENAI_API_KEY is not set");
if (!composioAPIKey) throw new Error("COMPOSIO_API_KEY is not set");
if (!composioUserID) throw new Error("COMPOSIO_USER_ID is not set");

const composio = new Composio({
  apiKey: composioAPIKey,
});

async function main() {
  // Create a tool router session for the user
  const session = await composio.create(composioUserID!, {
    toolkits: ["youtube"],
  });

  const mcpUrl = session.mcp.url;

  const mcpClient = await createMCPClient({
    transport: {
      type: "http",
      url: mcpUrl,
      headers: session.mcp.headers, // Authentication headers for the Composio MCP server
    },
  });

  const tools = await mcpClient.tools();

  let messages: ModelMessage[] = [];

  console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n");
  console.log(
    "Ask any questions related to youtube, like summarize my last 5 emails, send an email, etc... :)))\n",
  );

  const rl = readline.createInterface({
    input: process.stdin,
    output: process.stdout,
    prompt: "> ",
  });

  rl.prompt();

  rl.on("line", async (userInput: string) => {
    const trimmedInput = userInput.trim();

    if (["exit", "quit", "bye"].includes(trimmedInput.toLowerCase())) {
      console.log("\nGoodbye!");
      rl.close();
      process.exit(0);
    }

    if (!trimmedInput) {
      rl.prompt();
      return;
    }

    messages.push({ role: "user", content: trimmedInput });
    console.log("\nAgent is thinking...\n");

    try {
      const stream = streamText({
        model: openai("gpt-5"),
        messages,
        tools,
        toolChoice: "auto",
        stopWhen: stepCountIs(10),
        onStepFinish: (step) => {
          for (const toolCall of step.toolCalls) {
            console.log(`[Using tool: ${toolCall.toolName}]`);
          }
          if (step.toolCalls.length > 0) {
            console.log(""); // Add space after tool calls
          }
        },
      });

      for await (const chunk of stream.textStream) {
        process.stdout.write(chunk);
      }

      console.log("\n\n---\n");

      // Get final result for message history
      const response = await stream.response;
      if (response?.messages?.length) {
        messages.push(...response.messages);
      }
    } catch (error) {
      console.error("\nAn error occurred while talking to the agent:");
      console.error(error);
      console.log(
        "\nYou can try again or restart the app if it keeps happening.\n",
      );
    } finally {
      rl.prompt();
    }
  });

  rl.on("close", async () => {
    await mcpClient.close();
    console.log("\n👋 Session ended.");
    process.exit(0);
  });
}

main().catch((err) => {
  console.error("Fatal error:", err);
  process.exit(1);
});

Conclusion

You've successfully built a Youtube agent using the Vercel AI SDK with streaming capabilities! This implementation provides a powerful foundation for building AI applications with natural language interfaces and real-time feedback.

Key features of this implementation:

  • Real-time streaming responses for a better user experience with typewriter effect
  • Live tool execution feedback showing which tools are being used as the agent works
  • Dynamic tool loading through Composio's Tool Router with secure authentication
  • Multi-step tool execution with configurable step limits (up to 10 steps)
  • Comprehensive error handling for robust agent execution
  • Conversation history maintenance for context-aware responses

You can extend this further by adding custom error handling, implementing specific business logic, or integrating additional Composio toolkits to create multi-app workflows.

How to build Youtube MCP Agent with another framework

ChatGPT logo

ChatGPT

Use Youtube MCP with ChatGPT

Antigravity logo

Antigravity

Use Youtube MCP with Antigravity

OpenAI Agents SDK logo

OpenAI Agents SDK

Use Youtube MCP with OpenAI Agents SDK

Claude Agent SDK logo

Claude Agent SDK

Use Youtube MCP with Claude Agent SDK

Claude Code logo

Claude Code

Use Youtube MCP with Claude Code

Claude Cowork logo

Claude Cowork

Use Youtube MCP with Claude Cowork

Codex logo

Codex

Use Youtube MCP with Codex

Cursor logo

Cursor

Use Youtube MCP with Cursor

OpenClaw logo

OpenClaw

Use Youtube MCP with OpenClaw

Hermes logo

Hermes

Use Youtube MCP with Hermes

CLI logo

CLI

Use Youtube MCP with CLI

Google ADK logo

Google ADK

Use Youtube MCP with Google ADK

LangChain logo

LangChain

Use Youtube MCP with LangChain

Mastra AI logo

Mastra AI

Use Youtube MCP with Mastra AI

LlamaIndex logo

LlamaIndex

Use Youtube MCP with LlamaIndex

CrewAI logo

CrewAI

Use Youtube MCP with CrewAI

OpenCode logo

OpenCode

Use Youtube MCP with OpenCode

VS Code logo

VS Code

Use Youtube MCP with VS Code

FAQ

What are the differences in Tool Router MCP and Youtube MCP?

With a standalone Youtube MCP server, the agents and LLMs can only access a fixed set of Youtube tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Youtube and many other apps based on the task at hand, all through a single MCP endpoint.

Can I use Tool Router MCP with Vercel AI SDK v6?

Yes, you can. Vercel AI SDK v6 fully supports MCP integration. You get structured tool calling, message history handling, and model orchestration while Tool Router takes care of discovering and serving the right Youtube tools.

Can I manage the permissions and scopes for Youtube while using Tool Router?

Yes, absolutely. You can configure which Youtube scopes and actions are allowed when connecting your account to Composio. You can also bring your own OAuth credentials or API configuration so you keep full control over what the agent can do.

How safe is my data with Composio Tool Router?

All sensitive data such as tokens, keys, and configuration is fully encrypted at rest and in transit. Composio is SOC 2 Type 2 compliant and follows strict security practices so your Youtube data and credentials are handled as safely as possible.

Used by agents from

Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai
Context
Letta
glean
HubSpot
Agent.ai
Altera
DataStax
Entelligence
Rolai

Never worry about agent reliability

We handle tool reliability, observability, and security so you never have to second-guess an agent action.

Get started for freeGet a demo↗