How to integrate Cloudinary MCP with LangChain

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

30 min · no commitment · see it on your stack

Cloudinary logo
LangChain logo
divider

Introduction

This guide walks you through connecting Cloudinary to LangChain using the Composio tool router. By the end, you'll have a working Cloudinary agent that can create a new folder for event photos, delete derived assets with ids [123,456], set up upload preset with watermarking through natural language commands.

This guide will help you understand how to give your LangChain agent real control over a Cloudinary account through Composio's Cloudinary MCP server.

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

Also integrate Cloudinary with

TL;DR

Here's what you'll learn:
  • Get and set up your OpenAI and Composio API keys
  • Connect your Cloudinary project to Composio
  • Create a Tool Router MCP session for Cloudinary
  • Initialize an MCP client and retrieve Cloudinary tools
  • Build a LangChain agent that can interact with Cloudinary
  • Set up an interactive chat interface for testing

What is LangChain?

LangChain is a framework for developing applications powered by language models. It provides tools and abstractions for building agents that can reason, use tools, and maintain conversation context.

Key features include:

  • Agent Framework: Build agents that can use tools and make decisions
  • MCP Integration: Connect to external services through Model Context Protocol adapters
  • Memory Management: Maintain conversation history across interactions
  • Multi-Provider Support: Works with OpenAI, Anthropic, and other LLM providers

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

The Cloudinary MCP server is an implementation of the Model Context Protocol that connects your AI agent and assistants like Claude, Cursor, etc directly to your Cloudinary account. It provides structured and secure access to your digital asset management system, so your agent can perform actions like organizing folders, creating metadata fields, managing upload presets, and handling asset deletion on your behalf.

  • Automated folder and asset organization: Easily instruct your agent to create new asset folders or remove empty ones, keeping your Cloudinary library tidy and structured.
  • Metadata management: Let your agent create custom metadata fields or delete obsolete ones, extending and refining your asset tagging and search capabilities.
  • Preset and upload mapping creation: Have your agent set up upload presets with specific options or define dynamic folder mappings, automating consistent upload processes across your assets.
  • Resource and derived asset cleanup: Direct your agent to permanently delete assets by ID or remove unnecessary derived resources, ensuring your storage stays efficient and clutter-free.
  • Datasource entry management: Ask your agent to inactivate or delete specific datasource entries from metadata fields, keeping your metadata schema accurate and up to date.

Supported Tools & Triggers

Tools
Activate Live StreamTool to manually activate a Cloudinary live stream.
Create Asset Relations by Asset IDTool to add related assets by asset ID.
Create Asset Relations by Public IDTool to create relations between assets by public ID.
Create FolderTool to create a new asset folder.
Create Image from TextTool to create an image from text using Cloudinary's text generation API.
Create Live StreamTool to create a new live stream in Cloudinary.
Create Live Stream OutputTool to create a new live stream output configuration.
Create Metadata FieldTool to create a new metadata field definition.
Create Metadata RuleTool to create a new conditional metadata rule.
Create Multi-Resource AnimationTool to create an animated image, video, or PDF from a set of images.
Create SlideshowTool to create an auto-generated video slideshow from existing Cloudinary assets.
Create Streaming ProfileTool to create a new adaptive streaming profile in your Cloudinary account.
Create TransformationTool to create a new named transformation by assigning a custom name to a set of transformation parameters.
Create TriggerTool to create a new webhook trigger for a specified event type.
Create Upload MappingTool to create a new upload mapping folder and URL template.
Create Upload PresetTool to create a new upload preset.
Delete Asset Relations by Asset IDTool to delete asset relations by asset ID.
Delete Asset Relations by Public IDTool to delete asset relations by public ID.
Delete Derived ResourcesTool to delete derived assets.
Delete Metadata Field Datasource EntriesTool to delete datasource entries for a specified metadata field.
Delete FolderTool to delete an empty asset folder.
Delete Live StreamTool to delete a live stream from Cloudinary.
Delete Live Stream OutputTool to delete a live stream output from Cloudinary.
Delete Metadata FieldTool to delete a metadata field by external ID.
Delete Metadata RuleTool to delete a conditional metadata rule by its ID.
Delete Resources by Asset IDTool to delete resources by asset IDs.
Delete Resources by Public IDTool to delete Cloudinary resources by public ID, prefix, or all resources.
Delete Resources by TagsTool to delete Cloudinary assets by tag.
Delete Streaming ProfileTool to delete a custom streaming profile or revert a built-in profile to original settings.
Delete Transformation (v2)Tool to delete a named transformation from your Cloudinary account.
Delete TriggerTool to delete a trigger (webhook notification).
Delete Upload MappingTool to delete a folder upload mapping.
Delete Upload PresetTool to delete an upload preset from the account.
Destroy AssetTool to permanently destroy a Cloudinary asset/resource by public ID.
Destroy Asset by IDTool to delete an asset by its immutable asset ID.
Explicit Resource UpdateTool to update an existing asset and/or eagerly generate derived transformations using Cloudinary's Explicit API.
Explode Multi-Page ResourceTool to create derived images from multi-page files (PDF, PSD, TIFF, animated GIF) by exploding them into separate images.
Generate ArchiveTool to create an archive (ZIP or TGZ file) containing a set of assets from your Cloudinary environment.
Get Adaptive Streaming ProfilesTool to list adaptive streaming profiles.
Get Analysis Task StatusTool to get the status of an analysis task.
Get product environment config detailsTool to get product environment config details.
Get Live StreamTool to get details of a single live stream by ID.
Get Live Stream OutputTool to get details of a single live stream output.
Get Live Stream OutputsTool to get a list of live stream outputs.
Get Live StreamsTool to get a list of live streams from Cloudinary.
Get Metadata Field By IDTool to get a single metadata field definition by external ID.
Get Resource by Asset IDGet Resource by Asset ID
Get Resource by Public IDTool to get details of a single resource by public ID.
Get Resources by Asset FolderTool to list assets stored directly in a specified folder.
Get Resources by ContextTool to retrieve assets with a specified contextual metadata key/value.
Get Resources in ModerationTool to retrieve assets in a moderation queue by status.
Get Root FoldersTool to list all root folders in the product environment.
Get Streaming Profile DetailsTool to get details of a single streaming profile by name.
Get Resource TagsTool to list all tags used for a specified resource type.
Get TransformationTool to retrieve details of a specific transformation.
Get TransformationsTool to list all transformations (named and unnamed).
List Webhook TriggersTool to list all webhook triggers for event types in your environment.
Get Upload Mapping DetailsTool to retrieve details of a single upload mapping by folder.
Get Upload MappingsTool to list all upload mappings.
Get Upload PresetTool to retrieve details of a single upload preset by name.
Get UsageTool to get product environment usage details.
Get Video ViewsTool to get video analytics views from Cloudinary.
Idle Live StreamTool to manually idle a Cloudinary live stream.
List ImagesTool to list image assets from Cloudinary.
List Metadata FieldsTool to list all structured metadata fields defined in your Cloudinary product environment.
List Metadata RulesTool to retrieve all conditional metadata rules defined in your Cloudinary account.
List Raw FilesTool to list raw assets from Cloudinary.
List Resources by Asset IDsTool to retrieve multiple resources by their asset IDs.
List Resources by External IDsTool to retrieve resources by their external IDs.
List Resources by TagTool to list resources (assets) with a specified tag.
List Resources by TypeTool to retrieve resources (assets) by resource type and storage type.
List Resource TypesTool to list all available resource types in your Cloudinary product environment.
List Upload PresetsTool to list all upload presets configured in the account.
List Video AssetsTool to list video assets from Cloudinary.
Manage Context MetadataTool to add or remove contextual metadata on Cloudinary assets.
Order Metadata Field DatasourceTool to update ordering of a metadata field datasource.
Ping Cloudinary ServersTool to ping Cloudinary servers.
Publish ResourcesTool to publish Cloudinary assets by public IDs, prefix, or tag.
Rename or Move Resource Public IDTool to rename an asset's public ID using Cloudinary's rename endpoint.
Reorder Metadata FieldTool to change the position of a specific metadata field.
Reorder Metadata FieldsTool to reorder all metadata fields in the product environment.
Restore Metadata Field Datasource EntriesTool to restore previously deleted datasource entries for a metadata field.
Restore Deleted ResourcesTool to restore deleted Cloudinary resources by public IDs.
Restore Resources by Asset IDsTool to restore backed up assets by asset IDs.
Search AssetsTool to search and filter assets using powerful query expressions.
Search Datasource in Metadata FieldTool to search datasource values in a metadata field.
Search FoldersTool to search asset folders with filtering, sorting, and pagination.
Search All Metadata Field DatasourcesTool to search across all metadata field datasources.
Visual Search AssetsTool to find images in your asset library based on visual similarity or content.
Show FolderTool to list sub-folders within a specified folder.
Update Asset MetadataTool to populate or update metadata field values on one or more Cloudinary assets.
Update FolderTool to rename or move an existing asset folder.
Update Live StreamTool to update a live stream's configuration in Cloudinary.
Update Live Stream OutputTool to modify an existing live stream output configuration.
Update Metadata FieldTool to update a metadata field definition by external ID.
Update Metadata Field DatasourceTool to update the datasource (allowed values) for a metadata field.
Update Metadata RuleTool to update an existing conditional metadata rule.
Update Resource by Asset IDTool to update asset properties by asset ID in Cloudinary.
Update Resource by Public IDTool to update asset properties by public ID in Cloudinary.
Update Resource TagsTool to add, remove, replace, or remove all tags for one or more Cloudinary assets.
Update Streaming ProfileTool to modify an existing adaptive streaming profile's configuration.
Update Transformation (v2)Tool to update the definition of an existing named transformation.
Update TriggerTool to update the callback URL of an existing webhook trigger.
Update Upload MappingTool to update an existing upload mapping by changing its remote URL template.
Update Upload PresetTool to update an existing upload preset's configuration settings.
Upload AssetTool to upload media assets (images, videos, raw files) to Cloudinary.
Upload File ChunkTool to upload a single chunk of a large file to Cloudinary.
Upload File (Auto Detect)Tool to upload files with automatic resource type detection.

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

What is Composio SDK?

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

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

How the Composio SDK works

The Composio SDK follows a three-phase workflow:

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

Step-by-step Guide

Prerequisites

Before starting this tutorial, make sure you have:
  • Python 3.10 or higher installed on your system
  • A Composio account with an API key
  • An OpenAI API key
  • Basic familiarity with Python and async programming

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 dependencies

pip install composio-langchain langchain-mcp-adapters langchain python-dotenv

Install the required packages for LangChain with MCP support.

What's happening:

  • composio-langchain provides Composio integration for LangChain
  • langchain-mcp-adapters enables MCP client connections
  • langchain is the core agent framework
  • python-dotenv loads environment variables

Set up environment variables

bash
COMPOSIO_API_KEY=your_composio_api_key_here
COMPOSIO_USER_ID=your_composio_user_id_here
OPENAI_API_KEY=your_openai_api_key_here

Create a .env file in your project root.

What's happening:

  • COMPOSIO_API_KEY authenticates your requests to Composio's API
  • COMPOSIO_USER_ID identifies the user for session management
  • OPENAI_API_KEY enables access to OpenAI's language models

Import dependencies

from langchain_mcp_adapters.client import MultiServerMCPClient
from langchain.agents import create_agent
from dotenv import load_dotenv
from composio import Composio
import asyncio
import os

load_dotenv()
What's happening:
  • We're importing LangChain's MCP adapter and Composio SDK
  • The dotenv import loads environment variables from your .env file
  • This setup prepares the foundation for connecting LangChain with Cloudinary functionality through MCP

Initialize Composio client

async def main():
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))

    if not os.getenv("COMPOSIO_API_KEY"):
        raise ValueError("COMPOSIO_API_KEY is not set")
    if not os.getenv("COMPOSIO_USER_ID"):
        raise ValueError("COMPOSIO_USER_ID is not set")
What's happening:
  • We're loading the COMPOSIO_API_KEY from environment variables and validating it exists
  • Creating a Composio instance that will manage our connection to Cloudinary tools
  • Validating that COMPOSIO_USER_ID is also set before proceeding

Create a Tool Router session

# Create Tool Router session for Cloudinary
session = composio.create(
    user_id=os.getenv("COMPOSIO_USER_ID"),
    toolkits=['cloudinary']
)

url = session.mcp.url
What's happening:
  • We're creating a Tool Router session that gives your agent access to Cloudinary tools
  • The create method takes the user ID and specifies which toolkits should be available
  • The returned session.mcp.url is the MCP server URL that your agent will use
  • This approach allows the agent to dynamically load and use Cloudinary tools as needed

Configure the agent with the MCP URL

client = MultiServerMCPClient({
    "cloudinary-agent": {
        "transport": "streamable_http",
        "url": session.mcp.url,
        "headers": {
            "x-api-key": os.getenv("COMPOSIO_API_KEY")
        }
    }
})

tools = await client.get_tools()

agent = create_agent("gpt-5", tools)
What's happening:
  • We're creating a MultiServerMCPClient that connects to our Cloudinary MCP server via HTTP
  • The client is configured with a name and the URL from our Tool Router session
  • get_tools() retrieves all available Cloudinary tools that the agent can use
  • We're creating a LangChain agent using the GPT-5 model

Set up interactive chat interface

conversation_history = []

print("Chat started! Type 'exit' or 'quit' to end the conversation.\n")
print("Ask any Cloudinary related question or task to the agent.\n")

while True:
    user_input = input("You: ").strip()

    if user_input.lower() in ['exit', 'quit', 'bye']:
        print("\nGoodbye!")
        break

    if not user_input:
        continue

    conversation_history.append({"role": "user", "content": user_input})
    print("\nAgent is thinking...\n")

    response = await agent.ainvoke({"messages": conversation_history})
    conversation_history = response['messages']
    final_response = response['messages'][-1].content
    print(f"Agent: {final_response}\n")
What's happening:
  • We initialize an empty conversation_history list to maintain context across interactions
  • A while loop continuously accepts user input from the command line
  • When a user types a message, it's added to the conversation history and sent to the agent
  • The agent processes the request using the ainvoke() method with the full conversation history
  • Users can type 'exit', 'quit', or 'bye' to end the chat session gracefully

Run the application

if __name__ == "__main__":
    asyncio.run(main())
What's happening:
  • We call the main() function using asyncio.run() to start the application

Complete Code

Here's the complete code to get you started with Cloudinary and LangChain:

from langchain_mcp_adapters.client import MultiServerMCPClient
from langchain.agents import create_agent
from dotenv import load_dotenv
from composio import Composio
import asyncio
import os

load_dotenv()

async def main():
    composio = Composio(api_key=os.getenv("COMPOSIO_API_KEY"))
    
    if not os.getenv("COMPOSIO_API_KEY"):
        raise ValueError("COMPOSIO_API_KEY is not set")
    if not os.getenv("COMPOSIO_USER_ID"):
        raise ValueError("COMPOSIO_USER_ID is not set")
    
    session = composio.create(
        user_id=os.getenv("COMPOSIO_USER_ID"),
        toolkits=['cloudinary']
    )

    url = session.mcp.url
    
    client = MultiServerMCPClient({
        "cloudinary-agent": {
            "transport": "streamable_http",
            "url": url,
            "headers": {
                "x-api-key": os.getenv("COMPOSIO_API_KEY")
            }
        }
    })
    
    tools = await client.get_tools()
  
    agent = create_agent("gpt-5", tools)
    
    conversation_history = []
    
    print("Chat started! Type 'exit' or 'quit' to end the conversation.\n")
    print("Ask any Cloudinary related question or task to the agent.\n")
    
    while True:
        user_input = input("You: ").strip()
        
        if user_input.lower() in ['exit', 'quit', 'bye']:
            print("\nGoodbye!")
            break
        
        if not user_input:
            continue
        
        conversation_history.append({"role": "user", "content": user_input})
        print("\nAgent is thinking...\n")
        
        response = await agent.ainvoke({"messages": conversation_history})
        conversation_history = response['messages']
        final_response = response['messages'][-1].content
        print(f"Agent: {final_response}\n")

if __name__ == "__main__":
    asyncio.run(main())

Conclusion

You've successfully built a LangChain agent that can interact with Cloudinary through Composio's Tool Router.

Key features of this implementation:

  • Dynamic tool loading through Composio's Tool Router
  • Conversation history maintenance for context-aware responses
  • Async Python provides clean, efficient execution of agent workflows
You can extend this further by adding error handling, implementing specific business logic, or integrating additional Composio toolkits to create multi-app workflows.

How to build Cloudinary MCP Agent with another framework

OpenAI Agents SDK logo

OpenAI Agents SDK

Use Cloudinary MCP with OpenAI Agents SDK

Claude Agent SDK logo

Claude Agent SDK

Use Cloudinary MCP with Claude Agent SDK

Claude Code logo

Claude Code

Use Cloudinary MCP with Claude Code

Claude Cowork logo

Claude Cowork

Use Cloudinary MCP with Claude Cowork

Codex logo

Codex

Use Cloudinary MCP with Codex

OpenClaw logo

OpenClaw

Use Cloudinary MCP with OpenClaw

Hermes logo

Hermes

Use Cloudinary MCP with Hermes

CLI logo

CLI

Use Cloudinary MCP with CLI

Google ADK logo

Google ADK

Use Cloudinary MCP with Google ADK

Vercel AI SDK logo

Vercel AI SDK

Use Cloudinary MCP with Vercel AI SDK

Mastra AI logo

Mastra AI

Use Cloudinary MCP with Mastra AI

LlamaIndex logo

LlamaIndex

Use Cloudinary MCP with LlamaIndex

CrewAI logo

CrewAI

Use Cloudinary MCP with CrewAI

FAQ

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

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

Can I use Tool Router MCP with LangChain?

Yes, you can. LangChain 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 Cloudinary tools.

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

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

Used by agents from

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

Never worry about agent reliability

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

Get started for freeGet a demo↗