OpenAI API Responses Examples

This repository contains examples of using the OpenAI API's new Responses feature, which provides a more streamlined way to interact with OpenAI's models.

Setup

1. Make sure you have Python installed
2. Install the required packages:

   pip install openai rich pydantic requests
   

3. Create a .env file with your OpenAI API key:

   OPENAI_API_KEY=your_api_key_here
   

Examples

1. Basics (1_basics.py)

This script demonstrates the basic usage of the OpenAI Responses API:

• Creating simple responses with instructions
• Using the traditional messages array format
• Maintaining conversation context with previous_response_id
• Converting responses to dictionaries for inspection

2. Interactive Chat (2_chat.py)

A command-line chat interface built with the Rich library:

• Maintains conversation context across multiple exchanges
• Streams responses for a more interactive experience
• Displays token usage statistics
• Supports exiting the chat with the 'exit' command

3. Textual Chat (2a_textual_chat.py)

A Textual app for chatting with OpenAI's GPT models:

• Displays conversation history
• Shows token usage statistics
• Provides a simple text input for sending messages
• Streams responses as they are generated

4. Image Analysis (3_images.py)

Shows how to use the API to analyze images:

• Sends an image URL to the model
• Requests a description of the image content
• Uses the gpt-4o-mini model for efficient image analysis

5. Structured Outputs (4_structured_outputs.py)

Demonstrates how to get structured data from the API:

• Uses JSON Schema to define the expected output format
• Extracts structured information from natural language
• Shows integration with Pydantic for type validation
• Compares different methods for structured output

6. Web Search Integration (5_built_in_web_search.py)

Showcases the built-in web search capability:

• Configures the web search tool with user location
• Extracts and displays search results with annotations
• Uses Rich library to format and display the results in a visually appealing way
• Demonstrates how to parse the complex response structure

Usage

Run any example with Python:

python 1_basics.py

For the interactive chat example:

python 2_chat.py

Notes

• These examples use the latest OpenAI API features as of February 2025
• The Responses API provides a more streamlined interface compared to the older Chat Completions API
• The older Chat Completions API is still available and is still supported