Cloudinary has introduced its Model Context Protocol (MCP) Server, aiming to simplify the creation and management of images and video for online channels for users in both coding and non-coding roles.
The MCP Server is now available to all Cloudinary users, including those utilising the free plan. Through its integration with large language models (LLMs) such as Claude and Cursor, and compatibility with AI-platform Base44, the MCP Server enables these systems to interact with Cloudinary's image and video APIs using natural language, removing the requirement for manual coding.
For developers and non-technical users
Cloudinary's MCP Server targets three main user groups: developers working on complex visual media workflows for multi-channel or multi-country campaigns, non-technical app builders who wish to automate tasks such as image resizing and tagging, and general users who want to manage images and video assets without writing code.
In addition to the MCP Server, Cloudinary has released AI-ready documentation that provides LLMs with comprehensive information on the platform's APIs. This resource allows both developers and non-technical builders to access and leverage Cloudinary's capabilities quickly and with greater confidence, whether they are orchestrating advanced workflows or preparing assets for various online channels.
Cloudinary's integration with Base44 aims to support builders who want to incorporate Cloudinary's media transformation, optimisation, and delivery tools into their agent-driven workflows. This integration is intended to facilitate more streamlined development of visual-rich, scalable applications within agent-based environments.
Toolset and workflow automation
By utilising the combination of the new MCP Server and AI-ready documentation, users can gain access to a broad set of Cloudinary's features. This includes routine yet essential tasks such as bulk tagging, batch cropping, resizing, and overlaying images and video. Cloudinary states that these automated capabilities are designed to increase accuracy in media transformations while reducing the risk of model errors, often referred to as 'hallucinations' in AI workflows.
Through natural language interactions with LLMs, users can now connect with Cloudinary's APIs to complete such tasks without requiring additional integrations or coding expertise. According to the company, this can assist both technical teams looking to improve efficiency and cross-team collaboration, as well as newcomers to visual asset management aiming to build content-driven applications for digital channels.
Company statement
"Cloudinary was founded by developers for developers and MCP Server is just our latest commitment to ensuring developers and builders of all kinds have the tools they need to build the visual-first experiences and apps they dream about – faster and more easily than ever before," said Tal Lev-Ami, Co-founder and CTO, Cloudinary. "This new era of LLM-powered code generation underscores the importance of open, API-first platforms and tools designed to empower users to build within flexible yet trusted frameworks and models."
The company underscores its long-standing use of AI-driven development. Having integrated AI technologies since its inception, Cloudinary has supported businesses and developers in automating visual media processes to help ensure fast and tailored digital experiences for their audiences. Over the past year, the company has released multiple AI-based tools and enhancements, such as AI Vision and generative features like Generative Fill and Generative Upscale.
Community and access
Cloudinary is encouraging users to join its community and display their projects with the #BuiltWithCloudinary tag. The company also announced an MCP Server Challenge for developers and builders taking place this summer. All Cloudinary users will be invited to participate and will have the opportunity to win prizes.
The Cloudinary MCP Server is accessible to all developers and builders, and can be used with any Cloudinary account, including free plans.
