Getting Started with FastAPI Agents¶

This notebook walks you through the process of setting up a basic agent with Pydantic AI and FastAPI Agents for the first time step-by-step.

To run this notebook, make sure you have fastapi, fastapi-agents and pydantic-ai installed (e.g. pip install fastapi fastapi-agents pydantic-ai).

Create your Pydantic AI Agent¶

We're going to build a very simple agent to demo the combined power of these tools, a todo list manager!

Before running this, make sure OPENAI_API_KEY is set in your environment variables. If you need it, also set OPENAI_BASE_URL (e.g. if you are using models from GitHub Models).

To go to production we probably need to persist our todo lists somewhere like a database, but this is fine for a demo.

In [3]:

from typing import List
from pydantic_ai import Agent

todo_agent = Agent(
    "openai:gpt-4o",
    system_prompt="You are managing the user's todo list."
)

todos: List[str] = []

@todo_agent.tool
def list_todos(context):
    return f"{len(todos)} todos: {"".join(todos) if len(todos) > 0 else "empty"}"

@todo_agent.tool
def add_todo(context, description: str):
    todos.append(description)
    return f"added {description} to todos"

@todo_agent.tool
def delete_todo(context, description: str):
    todos.remove(description)
    return f"removed {description} from todos"

from typing import List from pydantic_ai import Agent todo_agent = Agent( "openai:gpt-4o", system_prompt="You are managing the user's todo list." ) todos: List[str] = [] @todo_agent.tool def list_todos(context): return f"{len(todos)} todos: {"".join(todos) if len(todos) > 0 else "empty"}" @todo_agent.tool def add_todo(context, description: str): todos.append(description) return f"added {description} to todos" @todo_agent.tool def delete_todo(context, description: str): todos.remove(description) return f"removed {description} from todos"

Test your agent¶

Let's make sure it works!

Since we are using Jupyter which uses asyncio for execution, we need to tell asyncio to nest the execution loop so we execute our Pydantic AI, which also uses asyncio. (You won't need this outside of Jupyter.)

In [ ]:

import nest_asyncio
nest_asyncio.apply()

result = todo_agent.run_sync("make up 3 todos and add them then tell me what do i need to do")
result.data

import nest_asyncio nest_asyncio.apply() result = todo_agent.run_sync("make up 3 todos and add them then tell me what do i need to do") result.data

Setup FastAPI¶

Start by creating your FastAPI instance. This gives us an app we can add our routes to and ultimately run.

In [5]:

from fastapi import FastAPI
app = FastAPI()

from fastapi import FastAPI app = FastAPI()

Setup FastAPI Agents¶

We need to configure our FastAPIAgents router. A router is just a place in FastAPI where routes can be defined.

Under the hood, FastAPIAgents is just a special kind of FastAPI APIRouter.

In [6]:

from fastapi_agents import FastAPIAgents

agents = FastAPIAgents(path_prefix="/agents")

from fastapi_agents import FastAPIAgents agents = FastAPIAgents(path_prefix="/agents")

Now we can register our todo agent. FastAPI Agents comes with a set of agent adapters for various frameworks. These adapters wrap the agent and convert it to FastAPI Agents standard interface.

We can then add the router to our FastAPI app.

In [7]:

from fastapi_agents.pydantic_ai import PydanticAIAgent

agents.register("todo", PydanticAIAgent(todo_agent))
app.include_router(agents)

from fastapi_agents.pydantic_ai import PydanticAIAgent agents.register("todo", PydanticAIAgent(todo_agent)) app.include_router(agents)

Serve our agent with FastAPI Agents!¶

Now we can start our app, serving our API with uvicorn.

Uvicorn will find an available port and start our API.

You can reach the docs at the endpoint /docs to interact with your agent through the API in your browser.

In [ ]:

import uvicorn

uvicorn.run(app)

import uvicorn uvicorn.run(app)