MCP transport (stdio, Streamable HTTP)
An MCP transport is the channel that carries MCP's JSON-RPC messages between client and server. The specification defines two: stdio, for a server run as a local subprocess, and Streamable HTTP, for a server reached over the network.
Also called: stdio, Streamable HTTP, SSE
By Divya Sharma · Checked Nov 15, 2026 · How we review
Every review in the MCP tools directory lists how a server connects, and that line names its transport. The choice decides where the server runs and how you set it up.
stdio
With stdio, the MCP client starts the server as a child process on the same machine. Messages go in through the server's standard input and come out of its standard output, one JSON message per line. Setup is a command and its arguments in the app's config. It needs no network port, but the server runs with your user's access to files and the network, so install only servers you trust. Local servers such as Filesystem usually use stdio.
Streamable HTTP
With Streamable HTTP, the server is an independent web service with a single endpoint, such as a URL ending in /mcp. The client sends each message as an HTTP POST. The server can reply with plain JSON or open a stream of Server-Sent Events (SSE) to send several messages, and the client can open a GET stream to receive messages the server starts. One server can handle many clients, which suits hosted services. Remote servers usually require sign-in, often through OAuth.
What happened to SSE
Early MCP had a separate HTTP+SSE transport. Streamable HTTP replaced the HTTP+SSE transport from the 2024-11-05 version of the specification. SSE still appears inside Streamable HTTP as an optional streaming format, and some clients keep the old transport for older servers.
Security notes
The specification says HTTP servers must check the Origin header, and local ones should listen only on localhost, to block DNS rebinding attacks from web pages.
Related terms
- MCP server
- An MCP server is a program that exposes tools, data resources or prompt templates to AI applications through the Model Context Protocol, so any MCP-compatible client can use them without custom integration code.
- MCP client
- An MCP client is the part of an AI application that holds one connection to one MCP server, negotiating capabilities and carrying requests, results and notifications between that server and the host app.
- Model Context Protocol
- The Model Context Protocol (MCP) is an open standard that defines how AI applications connect to external tools, data and prompts, using JSON-RPC messages between a host application, its clients and separate servers.
- OAuth for AI agents
- OAuth for AI agents is the use of the OAuth authorization standard to let an agent or MCP client act on a user's behalf with a limited, revocable access token, instead of holding the user's password or a broad API key.
- Streaming (LLM output)
- Streaming sends an LLM's output to the client piece by piece as tokens are generated, instead of waiting for the full response. Users see text appear within moments, even when the complete answer takes much longer.