Pydantic to OpenAI Strict Mode

Pydantic is the most popular Python library for data validation. Learn how to convert Pydantic models to OpenAI Structured Outputs compatible JSON Schema.

Why Pydantic for OpenAI?

Pydantic provides a clean, Pythonic way to define data structures. When working with OpenAI's Structured Outputs, you can:

• Define response schemas using familiar Python type hints
• Validate AI responses automatically
• Get IDE autocomplete and type checking
• Use model_json_schema() to generate JSON Schema

The Strict Mode Challenge

However, Pydantic's default model_json_schema() output is NOT compatible with OpenAI's strict mode because it lacks:

Solution: ConfigDict

Use Pydantic v2's ConfigDict with extra='forbid' to enforce strict mode:

from pydantic import BaseModel, ConfigDict

class MyResponse(BaseModel):
    model_config = ConfigDict(extra='forbid')
    
    name: str
    age: int
    active: bool

Using with OpenAI SDK

from openai import OpenAI
from pydantic import BaseModel, ConfigDict

class UserInfo(BaseModel):
    model_config = ConfigDict(extra='forbid')
    name: str
    email: str
    age: int

client = OpenAI()

response = client.chat.completions.create(
    model="gpt-4o-2024-08-06",
    messages=[{"role": "user", "content": "..."}],
    response_format={
        "type": "json_schema",
        "json_schema": {
            "name": "UserInfo",
            "strict": True,
            "schema": UserInfo.model_json_schema()
        }
    }
)