Quick Start

Installation

pip install llm-promptguard[openai]

Set your API key

export OPENAI_API_KEY=sk-...

Your first structured call

from promptguard import llm_call
from pydantic import BaseModel

class Sentiment(BaseModel):
    label: str
    confidence: float

result = llm_call(
    model="gpt-4o-mini",
    prompt="The product is fantastic and well-made",
    schema=Sentiment
)

print(result.data)
# Sentiment(label='positive', confidence=0.95)

Async usage

from promptguard import allm_call

result = await allm_call(
    model="gpt-4o",
    prompt="Summarize this document...",
    schema=Summary
)

Using different schema types

PromptGuard supports Pydantic models, TypedDicts, dataclasses, and raw JSON schemas:

from typing import TypedDict

class Person(TypedDict):
    name: str
    age: int

result = llm_call(model="gpt-4o", prompt="John is 30", schema=Person)

from dataclasses import dataclass

@dataclass
class Person:
    name: str
    age: int

result = llm_call(model="gpt-4o", prompt="John is 30", schema=Person)

Running regression tests

promptguard init    # set up in your project
promptguard test    # run snapshot tests

See the Testing guide for details.