# How to integrate Exa MCP with Vercel AI SDK v6 ```json { "title": "How to integrate Exa MCP with Vercel AI SDK v6", "toolkit": "Exa", "toolkit_slug": "exa", "framework": "Vercel AI SDK", "framework_slug": "ai-sdk", "url": "https://composio.dev/toolkits/exa/framework/ai-sdk", "markdown_url": "https://composio.dev/toolkits/exa/framework/ai-sdk.md", "updated_at": "2026-05-06T08:11:01.203Z" } ``` ## Introduction This guide walks you through connecting Exa to Vercel AI SDK v6 using the Composio tool router. By the end, you'll have a working Exa agent that can summarize recent news articles on ai safety, find similar research papers to this url, create a webset for quarterly sales data through natural language commands. This guide will help you understand how to give your Vercel AI SDK agent real control over a Exa account through Composio's Exa MCP server. Before we dive in, let's take a quick look at the key ideas and tools involved. ## Also integrate Exa with - [OpenAI Agents SDK](https://composio.dev/toolkits/exa/framework/open-ai-agents-sdk) - [Claude Agent SDK](https://composio.dev/toolkits/exa/framework/claude-agents-sdk) - [Claude Code](https://composio.dev/toolkits/exa/framework/claude-code) - [Claude Cowork](https://composio.dev/toolkits/exa/framework/claude-cowork) - [Codex](https://composio.dev/toolkits/exa/framework/codex) - [OpenClaw](https://composio.dev/toolkits/exa/framework/openclaw) - [Hermes](https://composio.dev/toolkits/exa/framework/hermes-agent) - [CLI](https://composio.dev/toolkits/exa/framework/cli) - [Google ADK](https://composio.dev/toolkits/exa/framework/google-adk) - [LangChain](https://composio.dev/toolkits/exa/framework/langchain) - [Mastra AI](https://composio.dev/toolkits/exa/framework/mastra-ai) - [LlamaIndex](https://composio.dev/toolkits/exa/framework/llama-index) - [CrewAI](https://composio.dev/toolkits/exa/framework/crew-ai) ## TL;DR Here's what you'll learn: - How to set up and configure a Vercel AI SDK agent with Exa integration - Using Composio's Tool Router to dynamically load and access Exa 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 Exa MCP server, and what's possible with it? The Exa MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Exa account. It provides structured and secure access to your Exa data platform, so your agent can perform actions like extracting answers from web data, running semantic searches, managing imports, and automating monitoring across your datasets. - Citation-backed question answering: Have your agent generate direct, source-cited answers or detailed summaries for your research questions using Exa’s advanced search. - Semantic similarity search: Quickly find web pages or documents that are semantically related to a given URL, complete with highlights or summaries for context. - Data import and webset management: Let your agent create, configure, or delete imports and websets to streamline data gathering and enrichment workflows. - Automated data monitoring: Schedule and manage monitors for websets to keep your data fresh and up-to-date with minimal manual intervention. - Event tracking and retrieval: Access a full history of system events or fetch details for specific events to stay on top of activity within your Exa environment. ## Supported Tools | Tool slug | Name | Description | |---|---|---| | `EXA_ANSWER` | Generate an answer | Generates a direct, citation-backed answer to a clear natural language question or topic using exa's search, adept at both specific answers and detailed summaries for open-ended queries. | | `EXA_CREATE_IMPORT` | Create Import | Tool to create a new import to upload data into a webset. use when you need to initialize an import before uploading the data file. | | `EXA_CREATE_MONITOR` | Create a Monitor | Tool to create a new monitor. use when you need to schedule automated updates for a webset without manual runs. | | `EXA_CREATE_WEBSET` | Create Webset | Tool to create a new webset with search, import, and enrichment setup. use when you need to configure and seed a webset in one call. | | `EXA_DELETE_IMPORT` | Delete import | Tool to delete an existing import. use when you need to permanently remove an import by its id. | | `EXA_DELETE_WEBSET` | Delete webset | Tool to delete a webset. use after confirming the webset id to permanently remove the webset and all its items. | | `EXA_FIND_SIMILAR` | Find similar | Finds web pages semantically similar to a given url using embeddings-based search, optionally retrieving full text, highlights, or summaries for results. | | `EXA_GET_CONTENTS_ACTION` | Get contents from URLs or document IDs | Retrieves configurable text and highlights from a list of exa document ids or publicly accessible urls. | | `EXA_GET_EVENT` | Get Event | Tool to get details of a specific event by its id. use when you have an event id and need its full details. | | `EXA_LIST_EVENTS` | List events | Tool to list all events that have occurred in the system. use when you need to paginate through the event history. | | `EXA_LIST_IMPORTS` | List imports | Tool to list all imports for the webset. use when you need to paginate through and monitor import jobs. | | `EXA_LIST_WEBHOOKS` | List webhooks | Tool to list all webhooks for websets. use when you need to view existing webhooks and paginate through results. | | `EXA_SEARCH` | Search | Performs a web search using the exa engine, useful for queries requiring advanced filtering, specific content categories, or ai-optimized prompting. | | `EXA_UPDATE_IMPORT` | Update import | Tool to update an import configuration by id. use when you need to modify an import's title or metadata. | ## Supported Triggers None listed. ## Creating MCP Server - Stand-alone vs Composio SDK The Exa MCP server is an implementation of the Model Context Protocol that connects your AI agent to Exa. It provides structured and secure access so your agent can perform Exa operations on your behalf through a secure, permission-based interface. With Composio's managed implementation, you don't have to create your own developer app. For production, if you're building an end product, we recommend using your own credentials. The managed server helps you prototype fast and go from 0-1 faster. ## Step-by-step Guide ### 1. Prerequisites Before you begin, make sure you have: - Node.js and npm installed - A Composio account with API key - An OpenAI API key ### 1. Getting API Keys for OpenAI and Composio OpenAI API Key - Go to the [OpenAI dashboard](https://platform.openai.com/settings/organization/api-keys) 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](https://dashboard.composio.dev?utm_source=toolkits&utm_medium=framework_docs). - Navigate to your API settings and generate a new API key. - Store this key securely as you'll need it for authentication. ### 2. Install required dependencies 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 ```bash npm install @ai-sdk/openai @ai-sdk/mcp @composio/core ai dotenv ``` ### 3. Set up environment variables 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 ```bash OPENAI_API_KEY=your_openai_api_key_here COMPOSIO_API_KEY=your_composio_api_key_here COMPOSIO_USER_ID=your_user_id_here ``` ### 4. Import required modules and validate environment 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 ```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, }); ``` ### 5. Create Tool Router session and initialize MCP client What's happening: - We're creating a Tool Router session that gives your agent access to Exa 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 Exa-related tools through the MCP protocol ```typescript async function main() { // Create a tool router session for the user const session = await composio.create(composioUserID!, { toolkits: ["exa"], }); const mcpUrl = session.mcp.url; ``` ### 6. Connect to MCP server and retrieve 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 Exa tools that the agent can use ```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(); ``` ### 7. Initialize conversation and CLI interface 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 ```typescript let messages: ModelMessage[] = []; console.log("Chat started! Type 'exit' or 'quit' to end the conversation.\n"); console.log( "Ask any questions related to exa, like summarize my last 5 emails, send an email, etc... :)))\n", ); const rl = readline.createInterface({ input: process.stdin, output: process.stdout, prompt: "> ", }); rl.prompt(); ``` ### 8. Handle user input and stream responses with real-time tool feedback 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 Exa 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 ```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); }); ``` ## Complete Code ```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: ["exa"], }); 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 exa, 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 Exa 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 Exa MCP Agent with another framework - [OpenAI Agents SDK](https://composio.dev/toolkits/exa/framework/open-ai-agents-sdk) - [Claude Agent SDK](https://composio.dev/toolkits/exa/framework/claude-agents-sdk) - [Claude Code](https://composio.dev/toolkits/exa/framework/claude-code) - [Claude Cowork](https://composio.dev/toolkits/exa/framework/claude-cowork) - [Codex](https://composio.dev/toolkits/exa/framework/codex) - [OpenClaw](https://composio.dev/toolkits/exa/framework/openclaw) - [Hermes](https://composio.dev/toolkits/exa/framework/hermes-agent) - [CLI](https://composio.dev/toolkits/exa/framework/cli) - [Google ADK](https://composio.dev/toolkits/exa/framework/google-adk) - [LangChain](https://composio.dev/toolkits/exa/framework/langchain) - [Mastra AI](https://composio.dev/toolkits/exa/framework/mastra-ai) - [LlamaIndex](https://composio.dev/toolkits/exa/framework/llama-index) - [CrewAI](https://composio.dev/toolkits/exa/framework/crew-ai) ## Related Toolkits - [Firecrawl](https://composio.dev/toolkits/firecrawl) - Firecrawl automates large-scale web crawling and data extraction. It helps organizations efficiently gather, index, and analyze content from online sources. - [Tavily](https://composio.dev/toolkits/tavily) - Tavily offers powerful search and data retrieval from documents, databases, and the web. It helps teams locate and filter information instantly, saving hours on research. - [Serpapi](https://composio.dev/toolkits/serpapi) - SerpApi is a real-time API for structured search engine results. It lets you automate SERP data collection, parsing, and analysis for SEO and research. - [Peopledatalabs](https://composio.dev/toolkits/peopledatalabs) - Peopledatalabs delivers B2B data enrichment and identity resolution APIs. Supercharge your apps with accurate, up-to-date business and contact data. - [Snowflake](https://composio.dev/toolkits/snowflake) - Snowflake is a cloud data warehouse built for elastic scaling, secure data sharing, and fast SQL analytics across major clouds. - [Posthog](https://composio.dev/toolkits/posthog) - PostHog is an open-source analytics platform for tracking user interactions and product metrics. It helps teams refine features, analyze funnels, and reduce churn with actionable insights. - [Amplitude](https://composio.dev/toolkits/amplitude) - Amplitude is a digital analytics platform for product and behavioral data insights. It helps teams analyze user journeys and make data-driven decisions quickly. - [Bright Data MCP](https://composio.dev/toolkits/brightdata_mcp) - Bright Data MCP is an AI-powered web scraping and data collection platform. Instantly access public web data in real time with advanced scraping tools. - [Browseai](https://composio.dev/toolkits/browseai) - Browseai is a web automation and data extraction platform that turns any website into an API. It's perfect for monitoring websites and retrieving structured data without manual scraping. - [ClickHouse](https://composio.dev/toolkits/clickhouse) - ClickHouse is an open-source, column-oriented database for real-time analytics and big data processing using SQL. Its lightning-fast query performance makes it ideal for handling large datasets and delivering instant insights. - [Coinmarketcal](https://composio.dev/toolkits/coinmarketcal) - CoinMarketCal is a community-powered crypto calendar for upcoming events, announcements, and releases. It helps traders track market-moving developments and stay ahead in the crypto space. - [Control d](https://composio.dev/toolkits/control_d) - Control d is a customizable DNS filtering and traffic redirection platform. It helps you manage internet access, enforce policies, and monitor usage across devices and networks. - [Databox](https://composio.dev/toolkits/databox) - Databox is a business analytics platform that connects your data from any tool and device. It helps you track KPIs, build dashboards, and discover actionable insights. - [Databricks](https://composio.dev/toolkits/databricks) - Databricks is a unified analytics platform for big data and AI on the lakehouse architecture. It empowers data teams to collaborate, analyze, and build scalable solutions efficiently. - [Datagma](https://composio.dev/toolkits/datagma) - Datagma delivers data intelligence and analytics for business growth and market discovery. Get actionable market insights and track competitors to inform your strategy. - [Delighted](https://composio.dev/toolkits/delighted) - Delighted is a customer feedback platform based on the Net Promoter System®. It helps you quickly gather, track, and act on customer sentiment. - [Dovetail](https://composio.dev/toolkits/dovetail) - Dovetail is a research analysis platform for transcript review and insight generation. It helps teams code interviews, analyze feedback, and create actionable research summaries. - [Dub](https://composio.dev/toolkits/dub) - Dub is a short link management platform with analytics and API access. Use it to easily create, manage, and track branded short links for your business. - [Elasticsearch](https://composio.dev/toolkits/elasticsearch) - Elasticsearch is a distributed, RESTful search and analytics engine for all types of data. It delivers fast, scalable search and powerful analytics across massive datasets. - [Fireflies](https://composio.dev/toolkits/fireflies) - Fireflies.ai is an AI-powered meeting assistant that records, transcribes, and analyzes voice conversations. It helps teams capture call notes automatically and search or summarize meetings effortlessly. ## Frequently Asked Questions ### What are the differences in Tool Router MCP and Exa MCP? With a standalone Exa MCP server, the agents and LLMs can only access a fixed set of Exa tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Exa 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 Exa tools. ### Can I manage the permissions and scopes for Exa while using Tool Router? Yes, absolutely. You can configure which Exa 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 Exa data and credentials are handled as safely as possible. --- [See all toolkits](https://composio.dev/toolkits) · [Composio docs](https://docs.composio.dev/llms.txt)