Official Docs

Chat Completions API Guide

How to call chat models, structure messages, and get reliable results.

Overview

The Chat Completions API allows you to build applications with GPT-3.5-turbo and GPT-4 models for various tasks including email drafting, code writing, document Q&A, conversational agents, language translation, and more. The API uses a message-based format that makes multi-turn conversations easy while remaining useful for single-turn tasks.

Key Points

Uses a series of messages as input with role-based structure

Supports system, user, and assistant message roles

System messages help set the behavior of the assistant

Conversations can be as short as 1 message or fill many pages

Designed for both single-turn and multi-turn conversations

Returns model-generated messages as structured output

Prerequisites

OpenAI API key and account

Understanding of message-based conversation patterns

Basic knowledge of API integration

Step-by-Step Guide

1Structure your conversation with appropriate message roles

2Start with a system message to set assistant behavior

3Add user messages for input and assistant messages for context

4Make API calls with the messages array

5Handle the response and extract the generated content

6Implement conversation state management for multi-turn chats