SIGN UPLOG IN
X (Twitter)LinkedInGitHubDiscordYouTube

How to integrate Parallel MCP with Vercel AI SDK v6

Connect Vercel AI SDK v6 to Parallel MCP. Find top articles on generative ai trends, summarize recent news about electric vehicles, and more using natural language, with authentication handled for you.

Get started for freeGet a demo
Parallel logoParallel
Api Key

Parallel is a Task API for automated, structured web research and data extraction. It transforms natural language queries into precise, schema-driven outputs for streamlined workflows.

42 Tools

Introduction

This guide walks you through connecting Parallel to Vercel AI SDK v6 using the Composio tool router. By the end, you'll have a working Parallel agent that can run research tasks, search and extract web content, find matching entities, set up monitors, and generate grounded answers through natural language commands.

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

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

Also integrate Parallel with

Claude CodeClaude CoworkOpenAI Agents SDKClaude Agent SDKCodexOpenClawHermesAtomic AgentGoogle ADKLangChainMastra AILlamaIndexCrewAIPydantic AIAutoGen

TL;DR

Here's what you'll learn:
  • How to set up and configure a Vercel AI SDK agent with Parallel integration
  • Using Composio's Tool Router to dynamically load and access Parallel 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 Parallel MCP server, and what's possible with it?

The Parallel MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Parallel account. It provides structured and secure access to advanced web research automation, so your agent can perform actions like launching batch research tasks, running semantic searches, monitoring task progress, and generating research suggestions on your behalf.

  • Automated web research task creation: Instantly create structured research tasks or batch multiple queries for parallel execution, saving time and effort.
  • Semantic search across multiple topics: Direct your agent to run parallel semantic searches and retrieve top-matching documents or data for several queries at once.
  • Real-time task group monitoring: Let your agent stream live updates about the progress, completion, or status of ongoing research task groups.
  • Context-driven research suggestions: Have the agent suggest the next best research tasks based on your project or intent, keeping your workflow efficient and on track.
  • Task group retrieval and management: Fetch detailed information about specific research task groups to review results or track progress seamlessly.

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

Step by step09 STEPS
1

Prerequisites

Before you begin, make sure you have:
  • Node.js and npm installed
  • A Composio account with API key
  • An OpenAI API key
2

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.
3

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
4

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
5

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
6

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: ["parallel"],
  });

  const mcpUrl = session.mcp.url;
What's happening:
  • We're creating a Tool Router session that gives your agent access to Parallel 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 Parallel-related tools through the MCP protocol
7

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 Parallel tools that the agent can use
8

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 parallel, 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
9

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 Parallel 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 Parallel 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: ["parallel"],
  });

  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 parallel, 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 Parallel 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.
TOOLS

Supported Tools

Every Parallel action and event your agent gets out of the box.

Add Enrichment to FindAll Run

Tool to add an enrichment to a FindAll run.

Add Runs to Task Group

Tool to initiate multiple task runs within a TaskGroup.

Cancel FindAll Run

Tool to cancel an active FindAll run by findall_id.

Create Chat Completions

Tool to get realtime chat completions from Parallel AI.

Create Monitor

Tool to create a web monitor that periodically runs the specified query.

Create Task Group

Tool to create a new task group.

Create Task Run

Tool to create and initiate a task run.

Delete Monitor

Tool to delete a monitor, stopping all future executions.

Extend FindAll Run

Tool to extend a FindAll run by adding additional matches to the current match limit.

Extract Content from URLs

Tool to extract relevant content from specific web URLs.

Extract Web Content

Extract excerpts or full content from known web URLs with Parallel's current stable Extract API.

Fetch Task Group Runs

Fetch Task Group run events from a resumable stream, optionally filtering by status and including inputs and outputs.

Start FindAll Run

Tool to start a FindAll run.

Generate Grounded Response

Generate a synchronous answer grounded in live web research with URL citation annotations.

Get FindAll Run Result

Tool to fetch the final (or latest available) FindAll candidates and result payload for a run.

Get FindAll Run Schema

Tool to retrieve the schema configuration of a FindAll run by findall_id.

Get Monitor

Retrieve a monitor's configuration and runtime state by ID using Parallel's current stable Monitor API.

Get Task Group

Retrieve a current stable-API snapshot of a Task Group's aggregate run counts and status.

Ingest FindAll Run

Tool to transform a natural language search objective into a structured FindAll specification.

List Current Monitors

List monitors from Parallel's current stable Monitor API, newest first, with optional type and status filters.

List Monitor Events

Tool to list events for a monitor from up to the last 300 event groups.

List Monitors

List monitors with their configuration and current state.

List Task Group Runs

Collect Task Group runs from the current stable API's resumable event stream, optionally including structured inputs and outputs.

List Task Run Events

Collect progress and state events from a Task Run using the current stable API.

Make Decisions

Answer 1-100 named predicate, choice, or score questions from shared state using current web evidence.

Retrieve Event Group

Tool to retrieve an event group for a monitor.

Retrieve FindAll Run Status

Tool to retrieve status and metadata for a FindAll run by findall_id.

Retrieve Memory

Retrieve relevant memories ranked against a query, or recent memories when the query is empty or omitted.

Retrieve Monitor

Tool to retrieve a specific monitor by ID.

Retrieve Task Group

Tool to retrieve details of a specific task group.

Retrieve Task Group Run

Tool to retrieve run status by run_id for a task group.

Retrieve Task Run

Tool to retrieve run status by run_id.

Retrieve Task Run Input

Tool to retrieve the input data of a specific task run by run_id.

Retrieve Task Run Result

Tool to retrieve the result of a task run by run_id, blocking until the run completes.

Parallel Search

Tool to perform parallel semantic search.

Search Web

Search the web with Parallel's current stable Search API and return ranked results with markdown excerpts.

Simulate Event

Tool to simulate sending an event for a monitor.

Stream FindAll Events

Collect FindAll run events, including candidate discovery, matching progress and run status, and return them as a list after the event stream ends.

Stream Task Group Events

Collect Task Group events, including group status and task run updates, and return them as a list after the event stream ends.

Stream Task Run Events

Collect Task Run events, including progress updates and state changes, and return them as a list after the event stream ends.

Suggest Task

Tool to suggest tasks based on user intent.

Update Monitor

Tool to update a monitor's configuration.

FRAMEWORKS

How to build Parallel MCP Agent with another framework

Claude Code logo

Claude Code

Use Parallel MCP with Claude Code

Claude Cowork logo

Claude Cowork

Use Parallel MCP with Claude Cowork

OpenAI Agents SDK logo

OpenAI Agents SDK

Use Parallel MCP with OpenAI Agents SDK

Claude Agent SDK logo

Claude Agent SDK

Use Parallel MCP with Claude Agent SDK

Codex logo

Codex

Use Parallel MCP with Codex

OpenClaw logo

OpenClaw

Use Parallel MCP with OpenClaw

Hermes logo

Hermes

Use Parallel MCP with Hermes

Atomic Agent logo

Atomic Agent

Use Parallel MCP with Atomic Agent

Google ADK logo

Google ADK

Use Parallel MCP with Google ADK

LangChain logo

LangChain

Use Parallel MCP with LangChain

Mastra AI logo

Mastra AI

Use Parallel MCP with Mastra AI

LlamaIndex logo

LlamaIndex

Use Parallel MCP with LlamaIndex

CrewAI logo

CrewAI

Use Parallel MCP with CrewAI

Pydantic AI logo

Pydantic AI

Use Parallel MCP with Pydantic AI

AutoGen logo

AutoGen

Use Parallel MCP with AutoGen

FAQ

Frequently asked questions

Start with Parallel.It takes 30 seconds.

Managed auth, hosted MCP servers, and every Parallel tool your agent needs.Free to start.

Start building