How to Use the Cloudinary MCP in LangChain

Q: How do I pass Cloudinary search results to a LangChain agent?

You map the output of searchmedialibrary directly into your agent's scratchpad. The agent reads the asset metadata and uses it to formulate the next tool call in the chain.

Q: Can a LangChain chain track Cloudinary bandwidth usage?

Yes, the agent calls getcloudinaryusagereport at the start of a chain to inspect your current consumption. You can write a routing function that branches your pipeline based on the remaining bandwidth.

Q: How does LangChain handle Cloudinary upload presets?

The agent executes listuploadpresets to find active configurations before preparing an upload. It then selects the correct preset name to pass to your upload handler in the next step.

Run multi-step media pipelines in LangChain by giving your agents direct, programmatic control over your Cloudinary assets.

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

MCP Servers - Free for Subscribers

Connect Cloudinary MCP to LangChain

Create your Vinkius account to connect Cloudinary to LangChain and route execution through our secure gateway. The platform manages server hosting, runtime updates, and security layers. Configuration requires no manual server provisioning.

GDPR Free for Subscribers

Setup Cloudinary with LangChain

Ask AI about this MCP

ChatGPT

Claude

Perplexity

Automate Cloudinary cleanup chains in LangChain

Your LangChain agent uses `list_media_resources` via this MCP Server to audit your cloud storage and identify stale assets. The agent evaluates the returned file metadata, decides which images are no longer needed, and immediately triggers `delete_media_resource` to purge them. Because every step runs inside a single reasoning loop, the output of your asset search feeds directly into the deletion tool. You can track this entire data flow and monitor tool latency inside your LangSmith dashboard.

Search Cloudinary libraries using this MCP Server

This MCP Server lets your LangChain agent run complex asset queries by invoking `search_media_library` dynamically based on user prompts. The agent parses natural language requests, translates them into Cloudinary search expressions, and executes the search. Once the agent gets the search results, it feeds the asset details into the next chain link to generate public URLs or analyze image tags. You get a transparent reasoning pipeline where the agent decides how to query your media catalog based on real-time feedback.

Monitor usage metrics in active chains

Your LangChain pipeline uses `get_cloudinary_usage_report` to check storage and bandwidth limits before starting heavy media operations. The agent inspects your remaining quota and halts execution or alerts your team if you are close to your monthly limits. Passing this usage data into a custom prompt template prevents failed uploads and keeps your automated workflows running within budget. Every API call and quota check is fully logged in your tracing tool for easy debugging.

Setup guide

Set up Cloudinary MCP in LangChain

Prerequisites

Python 3.10+ installed
langchain-mcp-adapters + langgraph packages
Active Vinkius subscription with a valid endpoint token

1

Install dependencies
Run pip install langchain-mcp-adapters langgraph langchain-openai. The MCP adapters package converts MCP tools into native LangChain BaseTool objects.
2

Connect via HTTP transport
Use MultiServerMCPClient with "transport": "http" pointing to your Vinkius endpoint. Replace [YOUR_TOKEN_HERE] with your token from cloud.vinkius.com.
3

Create a ReAct agent
Pass the discovered tools to create_react_agent() from LangGraph. The agent automatically routes Cloudinary tool calls through the MCP protocol.
4

Run with any LLM
Swap ChatOpenAI for ChatAnthropic, ChatGoogleGenerativeAI, or any LangChain-compatible model. The MCP tools work identically across all providers.

agent.py

from langchain_mcp_adapters.client import MultiServerMCPClient
from langgraph.prebuilt import create_react_agent
from langchain_openai import ChatOpenAI

async with MultiServerMCPClient({
    "cloudinary-mcp": {
        "transport": "http",
        "url": "https://edge.vinkius.com/[YOUR_TOKEN_HERE]/mcp",
    }
}) as client:
    tools = client.get_tools()

    agent = create_react_agent(
        ChatOpenAI(model="gpt-4o"),
        tools,
    )
    result = await agent.ainvoke({
        "messages": "List recent Cloudinary transactions"
    })
    print(result["messages"][-1].content)

Get your connection token →

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by Cloudinary. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

Why Choose Vinkius

Vinkius connects your tools to AI with real-time monitoring and automatic cost savings — all from one dashboard.

Connect Cloudinary now

Real-time monitoring

Live

visibility into every interaction

Connect your favorite tools to your AI and see exactly what's happening — every request, every response, in real time.

Built-in savings

60%

lower AI costs

Vinkius compresses data between your apps and your AI automatically. Lower bills every month — no configuration required.

Single dashboard

One

place for every integration

Every tool your AI connects to, managed from a single screen. One account, complete control.

Common questions about Cloudinary MCP in LangChain

You map the output of `search_media_library` directly into your agent's scratchpad. The agent reads the asset metadata and uses it to formulate the next tool call in the chain.

Yes, the agent calls `get_cloudinary_usage_report` at the start of a chain to inspect your current consumption. You can write a routing function that branches your pipeline based on the remaining bandwidth.

The agent executes `list_upload_presets` to find active configurations before preparing an upload. It then selects the correct preset name to pass to your upload handler in the next step.

Every tool call made by the MCP Server shows up in your LangSmith traces with exact inputs and outputs. You can see precisely when `get_media_resource_details` was called and how long the API took to respond.

This MCP Server only touches your Cloudinary media metadata, usage metrics, and asset tags. All communication runs locally inside a secure Vinkius sandbox, ensuring your private cloud credentials and asset lists never leak.

Use it with your favorite AI tools

Connect this server to Cursor, Claude, VS Code, and more.

OpenAI Agents SDK sdk-python

Google ADK sdk-python

Pydantic AI sdk-python

Vercel AI SDK sdk-typescript