Quick Start

This guide will help you get started with the OpenAIClient in datapizza-ai. For specialized topics, check out our detailed guides on multimodality, streaming and building chatbots.

Installation

First, make sure you have datapizza-ai installed:

pip install datapizza-ai

Basic Setup

from datapizza.clients.openai import OpenAIClient

# Initialize the client with your API key
client = OpenAIClient(
    api_key="your-openai-api-key",
    model="gpt-4o-mini",  # Default model
    system_prompt="You are a helpful assistant",  # Optional
    temperature=0.7  # Optional, controls randomness (0-2)
)

# Basic text response
response = client.invoke("What is the capital of France?")
print(response.text)
# Output: "The capital of France is Paris."

Core Methods

response = client.invoke(
    input="Explain quantum computing in simple terms",
    temperature=0.5,  # Override default temperature
    max_tokens=200,   # Limit response length
    system_prompt="You are a physics teacher"  # Override system prompt
)

print(response.text)
print(f"Tokens used: {response.completion_tokens_used}")

Async invoke

import asyncio

async def main():
    return await client.a_invoke(
        input="Explain quantum computing in simple terms",
        temperature=0.5,  # Override default temperature
        max_tokens=200,   # Limit response length
        system_prompt="You are a physics teacher"  # Override system prompt
    )

response = asyncio.run(main())

print(response.text)
print(f"Tokens used: {response.completion_tokens_used}")

Working with Memory

Memory allows you to maintain conversation context:

from datapizza.memory import Memory
from datapizza.type import ROLE, TextBlock

memory = Memory()

# First interaction
response1 = client.invoke("My name is Alice", memory=memory)
memory.add_turn(TextBlock(content="My name is Alice"), role=ROLE.USER)
memory.add_turn(response1.content, role=ROLE.ASSISTANT)

# Second interaction - the model remembers Alice
response2 = client.invoke("What's my name?", memory=memory)
print(response2.text)  # Should mention Alice

Token Management

Monitor your usage:

response = client.invoke("Explain AI")
print(f"Tokens used: {response.completion_tokens_used}")
print(f"Prompt token used: {response.prompt_tokens_used}")
print(f"Cached token used: {response.cached_tokens_used}")

That's it! You're ready to start building with the OpenAI client. Check out the specialized guides above for advanced features and patterns.

What's Next?

Now that you know the basics, explore our specialized guides:

📸 Multimodality Guide

Work with images, PDFs, and other media types for visual AI applications.

🌊 Streaming Guide

Build responsive applications with real-time text generation and streaming.

🛠️ Tools Guide

Extend AI capabilities by integrating external functions and tools.

📊 Structured Responses Guide

Work with strongly-typed outputs using JSON schemas and Pydantic models.

🤖 Chatbot Guide

Create sophisticated conversational AI with memory and context management.