Cloudinary MCP Server

Most developer tooling work still happens through a UI a human drives. Cloudinary MCP server moves it into the conversation instead. Connects assistants to a Cloudinary media library so they can upload files, run search and visual search, apply transformations, manage folders and tags.

The short version

Connects assistants to a Cloudinary media library so they can upload files, run search and visual search, apply transformations, manage folders and tags.

Getting it running

The server ships on npm as @cloudinary/asset-management-mcp, so your MCP client can launch it on demand — there is no separate build step. Add the server block to your client's configuration, restart it, and the tools register themselves.

The tools it exposes

The server publishes 20 tools. What each one is for:

• upload-asset — Upload an image, video or raw file from a local path or remote URL, setting public ID, folder and tags at ingest time
• search-assets — Run a structured expression search across the media library with sorting, aggregation and pagination
• visual-search-assets — Find assets by visual similarity or image content rather than by filename or tag
• transform-asset — Apply a Cloudinary transformation string to an existing asset via the explicit API, with optional CDN invalidation
• get-tx-reference — Return Cloudinary's full transformation reference so the model can compose valid transformation strings
• get-asset-details — Fetch the full metadata record for one asset by asset ID
• asset-update — Update an existing asset — tags, contextual metadata, moderation state and other mutable fields
• asset-rename — Change an asset's public ID, optionally overwriting an existing target
• delete-asset — Delete an asset by asset ID
• list-images — List image resources in the product environment with filtering and pagination
• list-videos — List video resources in the product environment
• list-files — List raw (non-image, non-video) files

What it needs from you

Configuration is passed through the environment: CLOUDINARY_CLOUD_NAME, CLOUDINARY_API_KEY, CLOUDINARY_API_SECRET, CLOUDINARY_URL, CLOUDINARY_MCP_APPS. Treat anything key-shaped as a real credential — scope it to the minimum the server needs, and rotate it if it ever lands in a shared config.

A Cloudinary account. The hosted endpoint authorises over OAuth; running the npm package locally needs Node.js 18+ and a cloud name, API key and API secret.

Things to watch

• It runs with your machine's permissions. That is convenient and also the reason to think about what you point it at before you approve a tool call.
• With 20 tools registered it takes up a noticeable share of the context window; turn it off in projects that never touch Cloudinary.
• Keep per-call confirmation enabled while you learn its behaviour; it is the cheapest safeguard you have.

How it compares

This sits in the developer tooling group, where several servers overlap in what they claim to do but differ sharply once you actually set them up. Cloudinary's toolset — upload-asset, search-assets, visual-search-assets and 17 more — is a fair guide to whether it matches your workflow. It is maintained by Cloudinary rather than a third party, which usually means auth handling and breaking changes are dealt with faster.

We check each listing at SyncDev against the project's documentation before it goes live — if something here drifts out of date, it is a bug worth reporting.