How to beginner · 3 min read

How to install Pydantic AI

Quick answer

Install pydantic-ai using pip install pydantic-ai in your Python environment. This package enables building AI agents with typed data models using Pydantic and supports integration with OpenAI and other LLMs.

PREREQUISITES

Python 3.8+
pip package manager
OpenAI API key (optional for usage)

Setup

Install pydantic-ai via pip in your Python environment. Ensure you have Python 3.8 or higher installed. Optionally, set your OpenAI API key as an environment variable for usage.

bash

pip install pydantic-ai

output

Collecting pydantic-ai
  Downloading pydantic_ai-0.1.0-py3-none-any.whl (15 kB)
Installing collected packages: pydantic-ai
Successfully installed pydantic-ai-0.1.0

Step by step

Here is a minimal example demonstrating how to use pydantic-ai with OpenAI's gpt-4o-mini model to create a typed AI agent that answers questions.

python

import os
from pydantic_ai import Agent
from pydantic import BaseModel

# Define a Pydantic model for structured output
class Result(BaseModel):
    answer: str
    confidence: float

# Initialize the agent with OpenAI model and system prompt
agent = Agent(
    "openai:gpt-4o-mini",
    result_type=Result,
    system_prompt="You are a helpful assistant."
)

# Run the agent synchronously with a question
result = agent.run_sync("What is Retrieval-Augmented Generation (RAG)?")

# Print the structured response
print(f"Answer: {result.data.answer}")
print(f"Confidence: {result.data.confidence}")

output

Answer: Retrieval-Augmented Generation (RAG) is a technique that combines retrieval of relevant documents with generative language models to produce accurate and context-aware responses.
Confidence: 0.95

Common variations

You can use pydantic-ai asynchronously with await agent.run() in async functions. It supports different LLM providers by changing the model string prefix (e.g., anthropic:claude-sonnet-4-5). You can also define custom Pydantic models for complex structured outputs.

python

import asyncio

async def async_example():
    result = await agent.run("Explain AI agent observability.")
    print(result.data.answer)

asyncio.run(async_example())

output

AI agent observability refers to the ability to monitor, trace, and analyze the behavior and outputs of AI agents to ensure reliability and improve performance.

Troubleshooting

If you see ModuleNotFoundError, ensure pydantic-ai is installed in the active Python environment.
If API calls fail, verify your OpenAI API key is set in the environment variable OPENAI_API_KEY.
For type validation errors, confirm your Pydantic models match the expected output structure.

✅

Key Takeaways

Use pip install pydantic-ai to install the package quickly.
pydantic-ai integrates typed Pydantic models with AI agents for structured outputs.
Set your OpenAI API key in OPENAI_API_KEY environment variable before running.
Supports synchronous and asynchronous usage patterns.
Customize output by defining your own Pydantic models.

Verified 2026-04 · gpt-4o-mini, claude-sonnet-4-5

Verify ↗