Cloudinary MCP Server

Official Cloudinary MCP servers for uploading, transforming, analyzing, and organizing media assets, plus configuring environment settings and workflow automation.

Design by Cloudinary OAuth2, API Key active

GitHub Docs npm: @cloudinary/asset-management-mcp

Overview

Cloudinary's MCP servers let AI agents like Claude, Cursor, Windsurf, and VS Code manage Cloudinary media operations through natural language. The official repo at cloudinary/mcp-servers ships five distinct servers that cover the major Cloudinary product surfaces: Asset Management, Environment Config, Structured Metadata, Analysis, and MediaFlows. Each can be run locally via npm/Docker or accessed through Cloudinary's hosted remote endpoints.

The Asset Management server handles uploads, search, folder/tag organization, transformations, and archive generation. The Environment Config server manages upload presets, named transformations, streaming profiles, webhooks, and upload mappings. Structured Metadata covers custom metadata fields, conditional rules, and validation. The Analysis server exposes AI-powered tools for tagging, moderation, captioning, object detection, watermark detection, and quality analysis. MediaFlows enables low-code workflow automation built from natural language.

Cloudinary hosts the remote variants under subdomains like asset-management.mcp.cloudinary.com/mcp with OAuth2 by default (and API key headers as an alternative). For local execution, each server is published as an npm package (e.g. @cloudinary/asset-management-mcp) that runs via npx. Official Docker images are also published. Some advanced features depend on the user's Cloudinary plan.

Tools

Tool	Description
`Asset Management`	Upload, search, organize, tag, transform, and archive images, videos, and raw files in Cloudinary.
`Environment Config`	Manage upload presets, named transformations, streaming profiles, webhook notifications, and upload mappings.
`Structured Metadata`	Define and administer structured metadata fields, values, conditional rules, and validation logic.
`Analysis`	AI-powered content analysis: automatic tagging, moderation, captioning, object detection, quality and watermark detection.
`MediaFlows`	Build and manage low-code automation workflows for media processing, moderation, and delivery using natural language.

Setup Guide

Prerequisites

A Cloudinary account (free tier works for basic operations; some features require paid plans)
Cloud name, API key, and API secret from the Cloudinary Console Dashboard
For local deployment: Node.js v18+ or Docker

Option 1: Remote Server (Recommended)

Cloudinary hosts each server. OAuth2 is supported by default, or you can pass credentials via headers.

{
  "mcpServers": {
    "cloudinary-asset-mgmt": {
      "url": "https://asset-management.mcp.cloudinary.com/mcp",
      "headers": {
        "cloudinary-url": "cloudinary://api_key:api_secret@cloud_name"
      }
    }
  }
}

Other remote endpoints:

https://environment-config.mcp.cloudinary.com/mcp
https://structured-metadata.mcp.cloudinary.com/mcp
https://analysis.mcp.cloudinary.com/sse (OAuth only)
https://mediaflows.mcp.cloudinary.com/v2/mcp

Option 2: Local Server via npx

{
  "mcpServers": {
    "cloudinary-asset-mgmt": {
      "command": "npx",
      "args": ["-y", "--package", "@cloudinary/asset-management-mcp", "--", "mcp", "start"],
      "env": {
        "CLOUDINARY_URL": "cloudinary://api_key:api_secret@cloud_name"
      }
    }
  }
}

Or use the individual variables CLOUDINARY_CLOUD_NAME, CLOUDINARY_API_KEY, and CLOUDINARY_API_SECRET.

Available npm packages:

@cloudinary/asset-management-mcp
@cloudinary/environment-config-mcp
@cloudinary/structured-metadata-mcp
@cloudinary/analysis

Option 3: Docker

Official images are published under the cloudinary namespace on Docker Hub for isolated, containerized deployment.

Use Cases

Upload product photos to Cloudinary, auto-tag them with the Analysis server, and organize them into folders by category
Bulk transform a folder of images (resize, format conversion, watermark) and generate signed delivery URLs for a marketing campaign
Run AI moderation on user-generated content uploads and flag assets containing inappropriate material or restricted objects
Define structured metadata fields (e.g. campaign, region, expiry date) and populate them across an asset library through natural language
Create MediaFlows automations that trigger on upload, run analysis, apply transformations, and route assets for approval

Example Prompts

"Upload all images in ./product-photos to the fall-2026 folder and tag them with season:fall and category:apparel."
"Search Cloudinary for all videos uploaded in the last 30 days that are over 50MB and list their public IDs."
"Run AI content moderation on the asset uploads/user_avatar_123 and tell me if it contains anything flagged."
"Create a new upload preset called signed-thumbnails that auto-generates a 300x300 webp version on upload."
"Build a MediaFlow that auto-captions any newly uploaded video and emails the result to the marketing team."

Pros

Official, maintained by Cloudinary across five distinct servers covering most of the platform
Multiple deployment options: hosted remote (with OAuth2), local npm, or Docker
Broad capability surface including upload, search, transformation, AI analysis, metadata, and workflow automation
Supports tool filtering by role (admin/builder/librarian) and regional endpoints (US/EU/AP)

Limitations

Capabilities are split across five separate servers, so users may need to install/configure multiple to cover all workflows
Some Analysis and MediaFlows features require paid Cloudinary plans
Large asset listings can exceed context limits; users must rely on filters and focused queries to avoid truncation

Alternatives

imgix MCP: community implementations for imgix image transformation API
AWS S3 MCP server: generic object storage if you only need upload/list/delete without media transforms
Filestack MCP: alternative media management API with community MCP wrappers