The fast, Pythonic way to build Model Context Protocol servers 🚀
You can now find FastMCP as part of the official Model Context Protocol Python SDK:
👉 github.com/modelcontextprotocol/python-sdk
Please note: this repository is no longer maintained.
The fast, Pythonic way to build MCP servers.
Model Context Protocol (MCP) servers are a new, standardized way to provide context and tools to your LLMs, and FastMCP makes building MCP servers simple and intuitive. Create tools, expose resources, and define prompts with clean, Pythonic code:
# demo.py
from fastmcp import FastMCP
mcp = FastMCP("Demo 🚀")
@mcp.tool()
def add(a: int, b: int) -> int:
"""Add two numbers"""
return a + b
That's it! Give Claude access to the server by running:
fastmcp install demo.py
FastMCP handles all the complex protocol details and server management, so you can focus on building great tools. It's designed to be high-level and Pythonic - in most cases, decorating a function is all you need.
(*emphasis on aims)
🚨 🚧 🏗️ FastMCP is under active development, as is the MCP specification itself. Core features are working but some advanced capabilities are still in progress.
We strongly recommend installing FastMCP with uv, as it is required for deploying servers:
uv pip install fastmcp
Note: on macOS, uv may need to be installed with Homebrew (brew install uv) in order to make it available to the Claude Desktop app.
Alternatively, to use the SDK without deploying, you may use pip:
pip install fastmcp
Let's create a simple MCP server that exposes a calculator tool and some data:
# server.py
from fastmcp import FastMCP
# Create an MCP server
mcp = FastMCP("Demo")
# Add an addition tool
@mcp.tool()
def add(a: int, b: int) -> int:
"""Add two numbers"""
return a + b
# Add a dynamic greeting resource
@mcp.resource("greeting://{name}")
def get_greeting(name: str) -> str:
"""Get a personalized greeting"""
return f"Hello, {name}!"
You can install this server in Claude Desktop and interact with it right away by running:
fastmcp install server.py
Alternatively, you can test it with the MCP Inspector:
fastmcp dev server.py
The Model Context Protocol (MCP) lets you build servers that expose data and functionality to LLM applications in a secure, standardized way. Think of it like a web API, but specifically designed for LLM interactions. MCP servers can:
There is a low-level Python SDK available for implementing the protocol directly, but FastMCP aims to make that easier by providing a high-level, Pythonic interface.
The FastMCP server is your core interface to the MCP protocol. It handles connection management, protocol compliance, and message routing:
from fastmcp import FastMCP
# Create a named server
mcp = FastMCP("My App")
# Specify dependencies for deployment and development
mcp = FastMCP("My App", dependencies=["pandas", "numpy"])
Resources are how you expose data to LLMs. They're similar to GET endpoints in a REST API - they provide data but shouldn't perform significant computation or have side effects. Some examples:
Resources can be static:
@mcp.resource("config://app")
def get_config() -> str:
"""Static configuration data"""
return "App configuration here"
Or dynamic with parameters (FastMCP automatically handles these as MCP templates):
@mcp.resource("users://{user_id}/profile")
def get_user_profile(user_id: str) -> str:
"""Dynamic user data"""
return f"Profile data for user {user_id}"
Tools let LLMs take actions through your server. Unlike resources, tools are expected to perform computation and have side effects. They're similar to POST endpoints in a REST API.
Simple calculation example:
@mcp.tool()
def calculate_bmi(weight_kg: float, height_m: float) -> float:
"""Calculate BMI given weight in kg and height in meters"""
return weight_kg / (height_m ** 2)
HTTP request example:
import httpx
@mcp.tool()
async def fetch_weather(city: str) -> str:
"""Fetch current weather for a city"""
async with httpx.AsyncClient() as client:
response = await client.get(
f"https://api.weather.com/{city}"
)
return response.text
Complex input handling example:
from pydantic import BaseModel, Field
from typing import Annotated
class ShrimpTank(BaseModel):
class Shrimp(BaseModel):
name: Annotated[str, Field(max_length=10)]
shrimp: list[Shrimp]
@mcp.tool()
def name_shrimp(
tank: ShrimpTank,
# You can use pydantic Field in function signatures for validation.
extra_names: Annotated[list[str], Field(max_length=10)],
) -> list[str]:
"""List all shrimp names in the tank"""
return [shrimp.name for shrimp in tank.shrimp] + extra_names
Prompts are reusable templates that help LLMs interact with your server effectively. They're like "best practices" encoded into your server. A prompt can be as simple as a string:
@mcp.prompt()
def review_code(code: str) -> str:
return f"Please review this code:\n\n{code}"
Or a more structured sequence of messages:
from fastmcp.prompts.base import UserMessage, AssistantMessage
@mcp.prompt()
def debug_error(error: str) -> list[Message]:
return [
UserMessage("I'm seeing this error:"),
UserMessage(error),
AssistantMessage("I'll help debug that. What have you tried so far?")
]
FastMCP provides an Image class that automatically handles image data in your server:
from fastmcp import FastMCP, Image
from PIL import Image as PILImage
@mcp.tool()
def create_thumbnail(image_path: str) -> Image:
"""Create a thumbnail from an image"""
img = PILImage.open(image_path)
img.thumbnail((100, 100))
# FastMCP automatically handles conversion and MIME types
return Image(data=img.tobytes(), format="png")
@mcp.tool()
def load_image(path: str) -> Image:
"""Load an image from disk"""
# FastMCP handles reading and format detection
return Image(path=path)
Images can be used as the result of both tools and resources.
The Context object gives your tools and resources access to MCP capabilities. To use it, add a parameter annotated with fastmcp.Context:
from fastmcp import FastMCP, Context
@mcp.tool()
async def long_task(files: list[str], ctx: Context) -> str:
"""Process multiple files with progress tracking"""
for i, file in enumerate(files):
ctx.info(f"Processing {file}")
await ctx.report_progress(i, len(files))
# Read another resource if needed
data = await ctx.read_resource(f"file://{file}")
return "Processing complete"
The Context object provides:
report_progress()debug(), info(), warning(), and error()read_resource()request_id and client_idThere are three main ways to use your FastMCP server, each suited for different stages of development:
The fastest way to test and debug your server is with the MCP Inspector:
fastmcp dev server.py
This launches a web interface where you can:
During development, you can:
--with:
fastmcp dev server.py --with pandas --with numpy
fastmcp dev server.py --with-editable .
Once your server is ready, install it in Claude Desktop to use it with Claude:
fastmcp install server.py
Your server will run in an isolated environment with:
mcp = FastMCP("My App", dependencies=["pandas", "numpy"])
--name:
fastmcp install server.py --name "My Analytics Server"
# Set variables individually
fastmcp install server.py -e API_KEY=abc123 -e DB_URL=postgres://...
# Or load from a .env file
fastmcp install server.py -f .env
For advanced scenarios like custom deployments or running without Claude, you can execute your server directly:
from fastmcp import FastMCP
mcp = FastMCP("My App")
if __name__ == "__main__":
mcp.run()
Run it with:
# Using the FastMCP CLI
fastmcp run server.py
# Or with Python/uv directly
python server.py
uv run python server.py
Note: When running directly, you are responsible for ensuring all dependencies are available in your environment. Any dependencies specified on the FastMCP instance are ignored.
Choose this method when you need:
All FastMCP commands will look for a server object called mcp, app, or server in your file. If you have a different object name or multiple servers in one file, use the syntax server.py:my_server:
# Using a standard name
fastmcp run server.py
# Using a custom name
fastmcp run server.py:my_custom_server
Here are a few examples of FastMCP servers. For more, see the examples/ directory.
A simple server demonstrating resources, tools, and prompts:
from fastmcp import FastMCP
mcp = FastMCP("Echo")
@mcp.resource("echo://{message}")
def echo_resource(message: str) -> str:
"""Echo a message as a resource"""
return f"Resource echo: {message}"
@mcp.tool()
def echo_tool(message: str) -> str:
"""Echo a message as a tool"""
return f"Tool echo: {message}"
@mcp.prompt()
def echo_prompt(message: str) -> str:
"""Create an echo prompt"""
return f"Please process this message: {message}"
A more complex example showing database integration:
from fastmcp import FastMCP
import sqlite3
mcp = FastMCP("SQLite Explorer")
@mcp.resource("schema://main")
def get_schema() -> str:
"""Provide the database schema as a resource"""
conn = sqlite3.connect("database.db")
schema = conn.execute(
"SELECT sql FROM sqlite_master WHERE type='table'"
).fetchall()
return "\n".join(sql[0] for sql in schema if sql[0])
@mcp.tool()
def query_data(sql: str) -> str:
"""Execute SQL queries safely"""
conn = sqlite3.connect("database.db")
try:
result = conn.execute(sql).fetchall()
return "\n".join(str(row) for row in result)
except Exception as e:
return f"Error: {str(e)}"
@mcp.prompt()
def analyze_table(table: str) -> str:
"""Create a prompt template for analyzing tables"""
return f"""Please analyze this database table:
Table: {table}
Schema:
{get_schema()}
What insights can you provide about the structure and relationships?"""
FastMCP requires Python 3.10+ and uv.
For development, we recommend installing FastMCP with development dependencies, which includes various utilities the maintainers find useful.
git clone https://github.com/jlowin/fastmcp.git
cd fastmcp
uv sync --frozen --extra dev
For running tests only (e.g., in CI), you only need the testing dependencies:
uv sync --frozen --extra tests
Please make sure to test any new functionality. Your tests should be simple and atomic and anticipate change rather than cement complex patterns.
Run tests from the root directory:
pytest -vv
FastMCP enforces a variety of required formats, which you can automatically enforce with pre-commit.
Install the pre-commit hooks:
pre-commit install
The hooks will now run on every commit (as well as on every PR). To run them manually:
pre-commit run --all-files
Fork the repository and create a new branch:
git checkout -b my-branch
Make your changes and commit them:
git add . && git commit -m "My changes"
Push your changes to your fork:
git push origin my-branch
Feel free to reach out in a GitHub issue or discussion if you have any questions!