Google cloud vision with Vercel AI SDK

How to integrate Google cloud vision MCP with Vercel AI SDK

Introduction TL;DR What is Vercel AI SDK What is the Google cloud vision MCP server Supported Tools & Triggers Creating MCP Server - Stand-alone vs Tool Router Step-by-step Guide Complete Code Conclusion How to build Google cloud vision MCP Agent with another framework Explore Other Toolkits FAQ

Introduction

This guide walks you through connecting Google cloud vision to Vercel AI SDK using the Composio tool router. By the end, you'll have a working Google cloud vision agent that can bulk import product images from gcs csv, list all vision ai service locations, create a new product for image recognition, delete an outdated product and its images through natural language commands.

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

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

TL;DR

Here's what you'll learn:

How to set up and configure a Vercel AI SDK agent with Google cloud vision integration
Using Composio's Tool Router to dynamically load and access Google cloud vision 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
Step Counting: Control multi-step tool execution
OpenAI Provider: Native integration with OpenAI models

What is the Google cloud vision MCP server, and what's possible with it?

The Google cloud vision MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Google Cloud Vision account. It provides structured and secure access to your image analysis resources, so your agent can perform actions like registering products, managing reference images, listing endpoints, and automating large-scale image operations on your behalf.

Product and reference image management: Easily create new products and add reference images for visual search, enabling your agent to organize and expand your vision datasets effortlessly.
Bulk import and product set operations: Let your agent import large numbers of reference images into product sets from Cloud Storage CSV files, streamlining dataset curation at scale.
Automated product cleanup and deletion: Direct your agent to purge unused or orphan products from your project, keeping your cloud resources tidy without manual effort.
Location and endpoint discovery: Quickly list available Vision AI service locations and existing IndexEndpoints, making it easy for your agent to select optimal regions and manage deployment targets.
Vision API operation tracking: Retrieve and review ongoing or past Vision API operations, so your agent can monitor processing jobs and ensure workflow transparency.

Supported Tools & Triggers

Tools

Create Vision ProductTool to create and return a new Product resource.

Create ReferenceImageTool to create a ReferenceImage under a product.

Delete ProductTool to permanently delete a Product and its reference images.

Get ProductTool to get information associated with a Product.

Get Product SetTool to get a ProductSet.

Import Product SetsTool to asynchronously import reference images into ProductSets from a CSV in GCS.

List IndexEndpointsTool to list IndexEndpoints in a project and location.

List LocationsTool to list available Vision AI service locations for a project.

List Vision API OperationsTool to list operations that match the specified filter.

Purge ProductsTool to asynchronously delete products in a ProductSet or orphan products.

Update ProductTool to update a Product's mutable fields: displayName, description, and productLabels.

Update Product SetTool to update a ProductSet resource.

Add Product to ProductSetTool to add a Product to a specified ProductSet.

Cancel Vision OperationTool to cancel a long-running Vision API operation.

Delete Vision API OperationTool to delete a long-running Vision API operation.

Delete Product SetTool to permanently delete a ProductSet.

Delete Reference ImageTool to permanently delete a reference image.

Get Vision API OperationTool to get the latest state of a long-running operation.

Get Reference ImageTool to get information associated with a ReferenceImage.

List Products in ProductSetTool to list Products in a specified ProductSet.

List ProjectsTool to list Google Cloud projects accessible by the authenticated user.

List Reference ImagesTool to list reference images for a product.

Remove Product from ProductSetTool to remove a Product from a specified ProductSet.

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

What is Tool Router?

Composio's Tool Router 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 Tool Router

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

How the Tool Router works

The Tool Router follows a three-phase workflow:

Discovery: Searches for tools matching your task and returns relevant toolkits with their details.
Authentication: Checks for active connections. If missing, creates an auth config and returns a connection URL via Auth Link.
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 { experimental_createMCPClient as 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: ["google_cloud_vision"],
  });

  const mcpUrl = session.mcp.url;

What's happening:

We're creating a Tool Router session that gives your agent access to Google cloud vision 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 Google cloud vision-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 Google cloud vision 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 google_cloud_vision, 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 Google cloud vision 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 Google cloud vision 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 { experimental_createMCPClient as 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: ["google_cloud_vision"],
  });

  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 google_cloud_vision, 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 Google cloud vision 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 Google cloud vision MCP Agent with another framework

OpenAI Agents SDK

Use Google cloud vision MCP with OpenAI Agents SDK

Claude Agent SDK

Use Google cloud vision MCP with Claude Agent SDK

Google ADK

Use Google cloud vision MCP with Google ADK

LangChain

Use Google cloud vision MCP with LangChain

Mastra AI

Use Google cloud vision MCP with Mastra AI

LlamaIndex

Use Google cloud vision MCP with LlamaIndex

CrewAI

Use Google cloud vision MCP with CrewAI

Explore Other Toolkits

Ai ml api

API Key

Ai ml api is a suite of AI/ML models for natural language and image tasks. It provides fast, scalable access to advanced AI capabilities for your apps and workflows.

Aivoov

API Key

Aivoov is an AI-powered text-to-speech platform offering 1,000+ voices in over 150 languages. Instantly turn written content into natural, human-like audio for any application.

All images ai

API Key

All-Images.ai is an AI-powered image generation and management platform. It helps you create, search, and organize images effortlessly with advanced AI capabilities.

TOOLKIT MARKETPLACE

FAQ

What are the differences in Tool Router MCP and Google cloud vision MCP?

With a standalone Google cloud vision MCP server, the agents and LLMs can only access a fixed set of Google cloud vision tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Google cloud vision 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?

Yes, you can. Vercel AI SDK 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 Google cloud vision tools.

Can I manage the permissions and scopes for Google cloud vision while using Tool Router?

Yes, absolutely. You can configure which Google cloud vision 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 Google cloud vision data and credentials are handled as safely as possible.

Used by agents from

Never worry about agent reliability

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

Book a call↗Get started for free

Harsha GaddipatiCo-founder, Slashy

Karan skipped his own birthday party to fix our critical issue. It was 10 pm and he diverted his Waymo to help us instead. This really sets the bar, shows you the commitment you need to have when users rely on your software.

Abhi AryaCo-founder, Opennote

A lot of students tell us that the moment their connected tools start talking to each other inside Opennote feels almost magical. The agent just knows them, and it has immensely helped in keeping new users on the platform.

Nirman DaveCEO, Zams

We chose Composio over Pipedream because it delivered depth where it mattered. It supported niche tools and tricky edge cases that other platforms simply ignored. Giving us confidence to scale without compromising.

Ryan YuFounder, Extra Thursday

As a solo builder, shipping fast is life or death. The only way I can outcompete incumbents is by outmanoeuvring them. Getting bogged down in the complexities of managing agent auth would have been a death sentence for Extra Thursday.

Tomisin JenrolaFounder & CEO, SwarmZero

Before partnering with Composio, adding tool integrations was a slow, resource-intensive process. Each integration could take weeks or months of engineering time, and maintaining them meant constantly keeping up with API changes.

Jerome LeclancheCo-Founder, Ingram Technologies

With hands-on help from their founder, we integrated Gmail and Google Drive in just 30 minutes. This level of personal support and commitment is exactly what startups should strive for.

Harsha GaddipatiCo-founder, Slashy

Abhi AryaCo-founder, Opennote

Nirman DaveCEO, Zams

Ryan YuFounder, Extra Thursday

Tomisin JenrolaFounder & CEO, SwarmZero

Jerome LeclancheCo-Founder, Ingram Technologies

With hands-on help from their founder, we integrated Gmail and Google Drive in just 30 minutes. This level of personal support and commitment is exactly what startups should strive for.

Harsha GaddipatiCo-founder, Slashy

Abhi AryaCo-founder, Opennote

Nirman DaveCEO, Zams

Ryan YuFounder, Extra Thursday

Tomisin JenrolaFounder & CEO, SwarmZero

Jerome LeclancheCo-Founder, Ingram Technologies

With hands-on help from their founder, we integrated Gmail and Google Drive in just 30 minutes. This level of personal support and commitment is exactly what startups should strive for.

How to integrate Google cloud vision MCP with Vercel AI SDK

Table of Contents

Introduction

TL;DR

What is Vercel AI SDK?

What is the Google cloud vision MCP server, and what's possible with it?

Supported Tools & Triggers

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

What is Tool Router?

How the Tool Router works

Step-by-step Guide

Prerequisites

Getting API Keys for OpenAI and Composio

Install required dependencies

Set up environment variables

Import required modules and validate environment

Create Tool Router session and initialize MCP client

Connect to MCP server and retrieve tools

Initialize conversation and CLI interface

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

Complete Code

Conclusion

How to build Google cloud vision MCP Agent with another framework

OpenAI Agents SDK

Claude Agent SDK

Google ADK

LangChain

Mastra AI

LlamaIndex

CrewAI

Explore Other Toolkits

Ai ml api

Aivoov

All images ai

FAQ

What are the differences in Tool Router MCP and Google cloud vision MCP?

Can I use Tool Router MCP with Vercel AI SDK?

Can I manage the permissions and scopes for Google cloud vision while using Tool Router?

How safe is my data with Composio Tool Router?

Used by agents from

Never worry about agent reliability