Anthropic, the company behind Claude 3.5, introduces Model Context Protocol (MCP), a new open source standard for efficiently connecting AI models to existing data systems, while facilitating their integration into varied business environments.
AI assistants, despite their increasing sophistication, often remain isolated from critical data. Tailored integrations between models and data systems fragment workflows, creating information silos that are costly to maintain and poorly scalable.
Connect AI assistants to the data they need
The Model Context Protocol responds to this challenge with a universal architecture.
“Whether you’re building an AI-powered IDE, enhancing a chat interface, or creating custom AI workflows, MCP provides a standardized way to connect LLMs with the context they need,” can we read on the MCP Github page.
This open standard allows developers to establish secure bidirectional connections between their data sources and LLMs. The architecture is simple: developers can either expose their data through MCP servers or create AI applications (MCP clients) that connect to these servers.
This contextualization protocol has three main components:
- protocol specification and software development kits (SDKs);
- local MCP server support in Claude Desktop applications;
- an open source repository of MCP servers.
To encourage adoption, Anthropic offers pre-built MCP servers for popular enterprise tools, such as Google Drive, Slack, GitHub, Postgres and Puppeteer.
Concrete use cases
Companies like Block and Apollo have already integrated MCP into their systems, while players such as Zed, Replit or Codeium rely on this protocol to enrich their platforms. These uses allow:
- Better context for complex tasks like coding, with more accurate code generation.
- A reduction in efforts linked to specific integrations, in favor of a standard approach.
Block, for example, sees MCP as a way to foster transparent and collaborative innovation. According to Dhanji R. Prasanna, CTO of Block:
“Open technologies like the Model Context Protocol are the bridges that connect AI to real-world applications, ensuring that innovation is accessible, transparent and rooted in collaboration. We are excited to partner with a protocol and use it to build agentic systems, which remove the burden of mechanics so people can focus on creating”.
How to get started with MCP?
Developers can explore MCP today with:
- Predefined MCP serversavailable via Claude Desktop ;
- A quick start guide to create a custom MCP server;
- Des contributions open source to available connectors and repositories.
Existing Claude for Work customers can begin testing MCP servers locally, connecting Claude to internal systems and data sets. Anthropic will soon provide development toolkits for deploying remote production MCP servers.
Related News :