MCP (Model Context Protocol)
An open standard that lets LLMs discover and call external tools, data, and services in a uniform way.
The Model Context Protocol, introduced by Anthropic in 2024 and now supported across OpenAI, Google, and the major agent frameworks, is a JSON-RPC standard for exposing tools, resources, and prompts to any MCP-aware LLM client.
Instead of writing bespoke integrations for every model provider, you build one MCP server per system (CRM, database, filesystem, ticketing) and it works with any compatible client. It is doing for AI tool-calling what USB did for peripherals.
For enterprises, MCP simplifies the security story: authentication, scopes, and audit logging live in one place per system, not scattered across every prompt.
Key points
- One server, many clients — no per-model integration work
- Standardized authentication and permission scopes
- Enables true tool marketplaces and reusable components
- Local (stdio) and remote (HTTP/SSE) transports supported
Common use cases
Frequently asked
Related terms
An LLM-driven system that plans, calls tools, and executes multi-step tasks autonomously.
A neural network trained on massive text corpora to understand and generate human-like language.
A structured pipeline where an LLM makes decisions at defined branch points, but the overall flow is deterministic.
Applying MCP (Model Context Protocol) in your business?
We help enterprises design, ship, and govern AI systems end-to-end.