When Anthropic unveiled Mannequin Context Protocol (MCP) final fall, it didn’t make a complete lot of headlines. Practically six months later, the brand new protocol is the speak of the city as one information infrastructure vendor after one other pronounces assist for the protocol, which was designed to simplify the mixing of AI fashions, information sources, and AI instruments.
The core concept behind MCP is pretty easy, based on Anthropic. As an alternative of constructing and sustaining particular person connectors for every information supply, AI software builders can make the most of a single, open protocol to make use of with all the connections.
“Consider MCP like a USB-C port for AI purposes,” Anthropic wrote in an introduction to MCP. “Simply as USB-C gives a standardized technique to join your units to varied peripherals and equipment, MCP gives a standardized technique to join AI fashions to completely different information sources and instruments.”
MCP makes use of a light-weight client-server structure. The server part of MCP resides on databases, file servers, or enterprise purposes, similar to a CRM system, whereas the MCP consumer resides within the AI software, similar to Anthropic’s Claude desktop or different software.
After a considerably gradual begin over the winter, MCP assist is choosing up in a giant means, significantly up to now month. Listed below are a few of the distributors now supporting MCP:
CData
One of many corporations launching assist for MCP is CData, the unheralded software program vendor that builds and helps most of the database connectors that database distributors license to do the arduous work of exposing their explicit database to the skin world utilizing commonplace interfaces and protocols, like ODBC and SQL, amongst others.
With its new MCP Servers providing, which is at present in beta, CDdata is supporting MCP as one of many transformation protocols it helps on the downstream aspect of its information adapters. It doesn’t but assist MCP in every of the 350-plus adapters that it builds and maintains, however the firm says that’s the plan.
“We’ve wrapped our connectors in an MCP server that exposes SQL-based instruments to LLMs,” CData CEO and Co-founder Amit Sharma explains in a current weblog publish. “This isn’t a one-off integration. It’s a general-purpose interface that enables LLMs to introspect enterprise information, write and execute SQL, and even carry out write operations–all on behalf of the consumer.”
Dremio
Dremio this week introduced assist for MCP with its information lakehouse platform. The corporate says the launch of Dremio MCP Server will allow “native integration” to AI fashions, similar to Anthropic’s Claude.
Dremio says the mix of its personal semantic layer with MCP assist will allow AI brokers to discover datasets, generate queries, and retrieve ruled information in actual time, whereas sustaining context throughout workflows.
Dremio’s implementation of MCP allows Claude to increase its reasoning capabilities on to a company’s information belongings, unlocking new potentialities for AI-powered insights whereas sustaining enterprise governance,” stated Mahesh Murag, an Anthropic product supervisor.
StarTree
Actual-time analytics database supplier StarTree can be getting in on the MCP motion. Final week, the firm introduced that its forthcoming assist for MCP will allow the StarTree database to serve information to “swarms of AI brokers” demanding entry to the newest, freshest information. The brand new launch additionally presents assist for internet hosting vector embeddings, which can bolster its prospects in serving information as a part of retrieval augmented era (RAG) workflows.
StarTree is an enterprise model of Apache Pinot, a real-time analytics database that’s recognized for its high-speed information ingestion and indexing capabilities in addition to the aptitude to serve giant numbers of SQL queries concurrently. Kishore Gopalakrishna, StarTree’s co-founder and CEO, says the brand new MCP and vector embedding capabilities will assist StarTree prospects push real-time information out to AI fashions and brokers.
“The subsequent wave of AI innovation will likely be pushed by real-time context—understanding what’s taking place now,” Gopalakrishna stated. “StarTree’s heritage as a real-time analytics basis completely enhances the place AI goes by delivering recent insights at scale. What’s altering is the shift from apps as the buyer to autonomous brokers.”
MindsDB
Database developer MindsDB is adopting MCP with each its open supply and enterprise platforms. The concept is to positions MindsDB “as a unified AI information hub that standardizes and optimizes how AI fashions entry information, dramatically simplifying synthetic intelligence deployment in advanced information environments,” the corporate stated in an April announcement.
MindsDB CEO Jorge Torres says his firm’s MCP assist will assist to reveal the proper piece of knowledge to the proper mannequin on the proper time with the proper safety controls, thereby serving to to “eradicate the information sprawl that’s posing a risk to enterprise AI innovation,” he stated.
Docker
Software program container supplier Docker final month introduced MCP assist in two new merchandise, together with Docker MCP Catalog and Docker MCP Toolkit. Docker MCP Catalog will present builders with a centralized technique to uncover verified and curated MCP instruments inside the Docker Hub, the corporate says, whereas Docker MCP Toolkit will let builders run, authenticate, and handle MCP instruments utilizing their commonplace Docker interface.
The brand new Docker choices will enable builders to construct AI and agentic purposes and deploy them as commonplace Docker containers. Docker says it can let “builders put AI to work with out reinventing their workflow.”
Docker can be constructing developer ecosystem for MCP instruments together with Elastic, Salesforce, and New Relic (amongst others) to do that. “Docker is positioning itself on the intersection of containerization and AI, the place pace, consistency, and safety are important,” stated Paul Nashawaty, Apply Lead and Principal Analyst at theCUBE Analysis.
Cloudflare
Cloud supplier Cloudflare final month introduced that it’s additionally supporting MCP. The corporate says its launching a distant MCP server that may allow builders to construct agentic AI purposes that may join over the Web and work together with providers, like electronic mail, “with out the necessity for a regionally hosted server.”
Its supporting MCP servers working as Sturdy Objects, which is a “particular kind of Cloudflare Employee that mix compute with storage, permitting you to construct stateful purposes in a serverless surroundings with out managing infrastructure.” On this method, MCP servers working on Cloudflare can retain context, which the corporate says will present a extra persistent and higher expertise for customers.
“Cloudflare presents a developer-friendly ecosystem for creating AI brokers that features a free tier for Sturdy Objects and serverless choices for AI inference,” explains
By offering a free tier for Sturdy Objects, Cloudflare is offering a developer-friendly ecosystem for creating AI brokers, says Kate Holterhoff, a senior analyst at RedMonk. “These low-cost, easy-to-use choices might empower extra companies to undertake and experiment with agentic AI,” she added.
Google Cloud
Google Cloud introduced assist for MCP throughout its Subsequent 2025 convention final month. Particularly, it’s supporting MCP in its Agent Backyard, a part of its AI providing the place prospects can browse ready-to-use samples of AI purposes after which begin constructing with them in its SDK.
Google Cloud additionally used the present to announce its personal Agent to Agent (A2A) protocol, which is geared at enabling brokers to name and connect with different brokers, versus AI fashions and instruments, which is the main target with MCP, based on Amin Vahdat, the corporate’s vp of ML, methods and cloud AI.
MinIO
MinIO, the developer of the S3-compatible object retailer of the identical time, can be supporting MCP. Final month, the corporate introduced assist the preview of MCP in AIStor, the enterprise model of its open supply object storage system.
The MCP functionality will allow brokers to “speak” to MinIO information by way of an MCP server, MinIO co-founder and co-CEO AB Periasamy stated. MinIO additionally introduced that it’s constructing administration capabilities into its MCP providing to stop unauthorized entry from MCP purchasers, in addition to introducing monitoring capabilities for MCP server.
Databricks
Databricks hasn’t formally introduced assist for MCP, and searches for MCP on its web site flip up empty. But it surely seems the corporate is supporting the protocol (or near supporting it) simply the identical.
In accordance with the outline of an upcoming session on Mosaic AI Agent Framework on the firm’s Knowledge + AI Summit (June 9-12 in San Francisco), MCP is a supported protocol in AI brokers which are callable from its AI Playground, whereas assist for MCP in Unity Catalog gives governance and safety for enterprise deployments.
There’s additionally an open supply Databricks MCP Server obtainable on GitHub that helps the Databricks API and permits LLMs and AI brokers to run SQL queries and in any other case entry Databricks environments.
There are a variety of different MCP implementations on the market amongst information platforms, clouds, database distributors, and others. As curiosity in agentic AI builds, builders will naturally gravitate in direction of instruments that permit them accomplish duties with the least quantity of value and complexity. It’s not completely clear that MCP will server as the information integration commonplace for agentic AI and GenAI in the long term. However within the brief time period, it’s undoubtedly off to an excellent begin.
Associated Objects:
Will Mannequin Context Protocol (MCP) Turn out to be the Customary for Agentic AI?
The Way forward for AI Brokers is Occasion-Pushed
LLMs and GenAI: When To Use Them