# How to integrate Nano nets MCP with Vercel AI SDK v6 ```json { "title": "How to integrate Nano nets MCP with Vercel AI SDK v6", "toolkit": "Nano nets", "toolkit_slug": "nano_nets", "framework": "Vercel AI SDK", "framework_slug": "ai-sdk", "url": "https://composio.dev/toolkits/nano_nets/framework/ai-sdk", "markdown_url": "https://composio.dev/toolkits/nano_nets/framework/ai-sdk.md", "updated_at": "2026-05-12T10:19:46.789Z" } ``` ## Introduction This guide walks you through connecting Nano nets to Vercel AI SDK v6 using the Composio tool router. By the end, you'll have a working Nano nets agent that can extract table data from recent invoices, upload new receipts for ocr model training, list all documents processed by a workflow through natural language commands. This guide will help you understand how to give your Vercel AI SDK agent real control over a Nano nets account through Composio's Nano nets MCP server. Before we dive in, let's take a quick look at the key ideas and tools involved. ## Also integrate Nano nets with - [OpenAI Agents SDK](https://composio.dev/toolkits/nano_nets/framework/open-ai-agents-sdk) - [Claude Agent SDK](https://composio.dev/toolkits/nano_nets/framework/claude-agents-sdk) - [Claude Code](https://composio.dev/toolkits/nano_nets/framework/claude-code) - [Claude Cowork](https://composio.dev/toolkits/nano_nets/framework/claude-cowork) - [Codex](https://composio.dev/toolkits/nano_nets/framework/codex) - [OpenClaw](https://composio.dev/toolkits/nano_nets/framework/openclaw) - [Hermes](https://composio.dev/toolkits/nano_nets/framework/hermes-agent) - [CLI](https://composio.dev/toolkits/nano_nets/framework/cli) - [Google ADK](https://composio.dev/toolkits/nano_nets/framework/google-adk) - [LangChain](https://composio.dev/toolkits/nano_nets/framework/langchain) - [Mastra AI](https://composio.dev/toolkits/nano_nets/framework/mastra-ai) - [LlamaIndex](https://composio.dev/toolkits/nano_nets/framework/llama-index) - [CrewAI](https://composio.dev/toolkits/nano_nets/framework/crew-ai) ## TL;DR Here's what you'll learn: - How to set up and configure a Vercel AI SDK agent with Nano nets integration - Using Composio's Tool Router to dynamically load and access Nano nets 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 Nano nets MCP server, and what's possible with it? The Nano nets MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Nano nets account. It provides structured and secure access to your intelligent document processing tools, so your agent can create, manage, and train OCR models, extract data from documents, and automate document workflows on your behalf. - Automated document data extraction: Let your agent process unstructured documents and pull out structured data using Nano nets' powerful AI-driven OCR models. - OCR model management: Easily create, update, and delete OCR models, allowing your agent to adjust to changing document types and business needs. - Workflow and document handling: Enable your agent to list, track, and manage documents within workflows, so you can monitor processing status and outcomes efficiently. - Training image uploads and model improvement: Have your agent upload new training images to OCR models, continually improving accuracy and adapting to new document formats. - Comprehensive model insights: Retrieve detailed information about your OCR models and their prediction files, empowering your agent to audit, debug, or optimize model performance as needed. ## Supported Tools | Tool slug | Name | Description | |---|---|---| | `NANO_NETS_CREATE_MODEL` | Create Model | Tool to create a new image classification or OCR model. Use when you need to initialize a model before uploading training images. Provide a list of categories/classes that the model should learn to identify or extract. | | `NANO_NETS_DELETE_MODEL` | Delete OCR Model | Permanently deletes an OCR model from Nanonets. Use this action when you need to remove a trained model that is no longer needed. This action is irreversible - once deleted, the model and all its training data cannot be recovered. Prerequisites: Obtain the model_id from the 'Get all OCR models' action first. | | `NANO_NETS_GET_ALL_MODELS` | Get All Models | Retrieves all models (OCR and Image Classification) in the user's NanoNets account. Returns model details including ID, type, status, accuracy, and extractable fields/categories. Use to discover available models before performing predictions or training operations. | | `NANO_NETS_GET_ALL_PREDICTION_FILES` | Get All Prediction Files | Retrieve all prediction files (OCR results) for a NanoNets model. Use this tool to: - List all documents/images that have been processed by an OCR model - Get prediction results including extracted text and field values - Access file URLs and processing status for each prediction The response includes prediction labels with extracted text, confidence scores, and bounding box coordinates for each processed file. | | `NANO_NETS_GET_MODEL_DETAILS` | Get OCR Model Details | Tool to retrieve details of an OCR model. Use when you need full metadata of a model by its ID. | | `NANO_NETS_GET_TRAINING_IMAGES` | Get OCR Training Images | Tool to retrieve training images for an OCR model. Use when you need to page through images associated with a model before training or analysis. | | `NANO_NETS_GET_WORKFLOWS` | Get Workflows | Tool to retrieve a list of all workflows in your Nanonets account. Use when you need to inventory or inspect all configured workflows. | | `NANO_NETS_LIST_DOCUMENTS` | List Workflow Documents | Retrieve a paginated list of documents processed by a NanoNets workflow. Returns document metadata including processing status, upload timestamp, verification status, and page details. Use this to monitor document processing progress or access extracted data from previously uploaded documents. | | `NANO_NETS_UPDATE_MODEL` | Update Model AI Guidelines | Update AI Agent guidelines for an OCR model. Sets instructions for how the AI should handle field and table predictions. Only works for Instant Learning models. Use this to customize extraction behavior for specific document types. | | `NANO_NETS_UPLOAD_TRAINING_IMAGES_BY_FILE` | Upload Training Images by File | Tool to upload a training image file to a specified OCR model. Use when adding a local image file to train the model. Supported file formats include PNG, JPEG, and PDF. | | `NANO_NETS_UPLOAD_TRAINING_IMAGES_BY_URL` | Upload Training Images by URL | Tool to upload training images by URL to a specified OCR model. Use when adding URLs of images to a model for training purposes. | ## Supported Triggers None listed. ## Creating MCP Server - Stand-alone vs Composio SDK The Nano nets MCP server is an implementation of the Model Context Protocol that connects your AI agent to Nano nets. It provides structured and secure access so your agent can perform Nano nets 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 Nano nets 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 Nano nets-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: ["nano_nets"], }); 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 Nano nets 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 nano_nets, 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 Nano nets 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("\n๐Ÿ‘‹ 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: ["nano_nets"], }); 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 nano_nets, 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("\n๐Ÿ‘‹ Session ended."); process.exit(0); }); } main().catch((err) => { console.error("Fatal error:", err); process.exit(1); }); ``` ## Conclusion You've successfully built a Nano nets 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 Nano nets MCP Agent with another framework - [OpenAI Agents SDK](https://composio.dev/toolkits/nano_nets/framework/open-ai-agents-sdk) - [Claude Agent SDK](https://composio.dev/toolkits/nano_nets/framework/claude-agents-sdk) - [Claude Code](https://composio.dev/toolkits/nano_nets/framework/claude-code) - [Claude Cowork](https://composio.dev/toolkits/nano_nets/framework/claude-cowork) - [Codex](https://composio.dev/toolkits/nano_nets/framework/codex) - [OpenClaw](https://composio.dev/toolkits/nano_nets/framework/openclaw) - [Hermes](https://composio.dev/toolkits/nano_nets/framework/hermes-agent) - [CLI](https://composio.dev/toolkits/nano_nets/framework/cli) - [Google ADK](https://composio.dev/toolkits/nano_nets/framework/google-adk) - [LangChain](https://composio.dev/toolkits/nano_nets/framework/langchain) - [Mastra AI](https://composio.dev/toolkits/nano_nets/framework/mastra-ai) - [LlamaIndex](https://composio.dev/toolkits/nano_nets/framework/llama-index) - [CrewAI](https://composio.dev/toolkits/nano_nets/framework/crew-ai) ## Related Toolkits - [Google Drive](https://composio.dev/toolkits/googledrive) - Google Drive is a cloud storage platform for uploading, sharing, and collaborating on files. It's perfect for keeping your documents accessible and organized across devices. - [Google Docs](https://composio.dev/toolkits/googledocs) - Google Docs is a cloud-based word processor that enables document creation and real-time collaboration. Its seamless sharing and version history make team editing and content management a breeze. - [Google Super](https://composio.dev/toolkits/googlesuper) - Google Super is an all-in-one suite combining Gmail, Drive, Calendar, Sheets, Analytics, and more. It gives you a unified platform to manage your digital life, boosting productivity and organization. - [Affinda](https://composio.dev/toolkits/affinda) - Affinda is an AI-powered document processing platform that automates data extraction from resumes, invoices, and more. It streamlines document-heavy workflows by turning files into structured, actionable data. - [Agility cms](https://composio.dev/toolkits/agility_cms) - Agility CMS is a headless content management system for building and managing digital experiences across platforms. It lets teams update content quickly and deliver omnichannel experiences with ease. - [Algodocs](https://composio.dev/toolkits/algodocs) - Algodocs is an AI-powered platform that automates data extraction from business documents. It delivers fast, secure, and accurate processing without templates or manual training. - [Api2pdf](https://composio.dev/toolkits/api2pdf) - Api2Pdf is a REST API for generating PDFs from HTML, URLs, and documents using powerful engines like wkhtmltopdf and Headless Chrome. It streamlines document conversion and automation for developers and businesses. - [Aryn](https://composio.dev/toolkits/aryn) - Aryn is an AI-powered platform for parsing, extracting, and analyzing data from unstructured documents. Use it to automate document processing and unlock actionable insights from your files. - [Boldsign](https://composio.dev/toolkits/boldsign) - Boldsign is a digital eSignature platform for sending, signing, and tracking documents online. Organizations use it to automate agreements and manage legally binding workflows efficiently. - [Boloforms](https://composio.dev/toolkits/boloforms) - BoloForms is an eSignature platform built for small businesses, offering unlimited signatures, templates, and forms. It simplifies digital document signing and team collaboration at a predictable, fixed price. - [Box](https://composio.dev/toolkits/box) - Box is a cloud content management and file sharing platform for businesses. It helps teams securely store, organize, and collaborate on files from anywhere. - [Carbone](https://composio.dev/toolkits/carbone) - Carbone is a blazing-fast report generator that turns JSON data into PDFs, Word docs, spreadsheets, and more using flexible templates. It lets you automate document creation at scale with minimal code. - [Castingwords](https://composio.dev/toolkits/castingwords) - CastingWords is a transcription service specializing in human-powered, accurate transcripts via a simple API. Get seamless audio-to-text conversion for interviews, meetings, podcasts, and more. - [Cloudconvert](https://composio.dev/toolkits/cloudconvert) - CloudConvert is a powerful file conversion service supporting over 200 file formats. It streamlines converting, compressing, and managing documents, media, and more, all in one place. - [Cloudlayer](https://composio.dev/toolkits/cloudlayer) - Cloudlayer is a document and asset generation service for creating PDFs and images via API or SDKs. It lets you automate high-quality doc creation, saving dev time and reducing manual work. - [Cloudpress](https://composio.dev/toolkits/cloudpress) - Cloudpress is a content export tool for Google Docs and Notion. It automates publishing to your favorite Content Management Systems. - [Contentful graphql](https://composio.dev/toolkits/contentful_graphql) - Contentful graphql is a content delivery API that lets you access Contentful data using GraphQL queries. It gives you efficient, flexible ways to fetch and manage structured content for any digital project. - [Conversion tools](https://composio.dev/toolkits/conversion_tools) - Conversion Tools is an online service for converting documents between formats such as PDF, Word, Excel, XML, and CSV. It lets you automate complex document workflows with just a few clicks. - [Convertapi](https://composio.dev/toolkits/convertapi) - ConvertAPI is a robust file conversion service for documents, images, and spreadsheets. It streamlines programmatic format changes and lets developers automate complex workflows with a single API. - [Craftmypdf](https://composio.dev/toolkits/craftmypdf) - CraftMyPDF is a web-based service for designing and generating PDFs with templates and live data. It streamlines document creation by automating personalized PDFs at scale. ## Frequently Asked Questions ### What are the differences in Tool Router MCP and Nano nets MCP? With a standalone Nano nets MCP server, the agents and LLMs can only access a fixed set of Nano nets tools tied to that server. However, with the Composio Tool Router, agents can dynamically load tools from Nano nets 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 Nano nets tools. ### Can I manage the permissions and scopes for Nano nets while using Tool Router? Yes, absolutely. You can configure which Nano nets 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 Nano nets data and credentials are handled as safely as possible. --- [See all toolkits](https://composio.dev/toolkits) ยท [Composio docs](https://docs.composio.dev/llms.txt)