Installation

Install the OpenAI Python SDK and configure your environment

This guide will help you install the OpenAI Python SDK and set up your development environment.

Requirements

Before installing the SDK, ensure you have:

• Python 3.9 or higher - Check your version with python --version
• pip - Python’s package installer (comes with Python)
• OpenAI API key - Get one here

Basic Installation

Install the OpenAI SDK from PyPI using pip:

pip install openai

This installs the core SDK with all essential dependencies:

• httpx - HTTP client for API requests
• pydantic - Data validation and type definitions
• typing-extensions - Type hint extensions
• anyio - Async I/O support
• distro - OS distribution detection
• tqdm - Progress bar support
• jiter - Fast JSON parsing

Optional Dependencies

The SDK provides several optional dependency groups for extended functionality:

Realtime API

For low-latency WebSocket connections with the Realtime API:

pip install openai[realtime]

This adds:

• websockets (>= 13, < 16) - WebSocket client for realtime connections

The Realtime API enables multi-modal conversational experiences with text and audio support.

Data Libraries

For working with data analysis and pandas DataFrames:

pip install openai[datalib]

This adds:

• numpy (>= 1) - Numerical computing
• pandas (>= 1.2.3) - Data manipulation
• pandas-stubs (>= 1.1.0.11) - Type stubs for pandas

Voice Helpers

For audio input/output with the Realtime API:

pip install openai[voice_helpers]

This adds:

• sounddevice (>= 0.5.1) - Audio I/O
• numpy (>= 2.0.2) - Audio data processing

Voice helpers are particularly useful when building interactive voice applications with the Realtime API.

Aiohttp (Async Performance)

For improved async performance with aiohttp as the HTTP backend:

pip install openai[aiohttp]

This adds:

• aiohttp - Alternative async HTTP client
• httpx_aiohttp (>= 0.1.9) - Aiohttp integration for httpx

To use aiohttp, instantiate the async client with DefaultAioHttpClient():

import asyncio
from openai import AsyncOpenAI, DefaultAioHttpClient

async def main():
    async with AsyncOpenAI(
        http_client=DefaultAioHttpClient(),
    ) as client:
        response = await client.responses.create(
            model="gpt-5.2",
            input="Say this is a test"
        )
        print(response.output_text)

asyncio.run(main())

Install Multiple Extensions

You can combine multiple optional dependencies:

Realtime + Voice

pip install openai[realtime,voice_helpers]

pip install openai[realtime,datalib,voice_helpers,aiohttp]

Environment Setup

1. Get your API key

   Sign up for an OpenAI account and get your API key from the API keys page.

2. Set environment variable

   Set the OPENAI_API_KEY environment variable:

   macOS/Linux

   export OPENAI_API_KEY="sk-proj-..."
   

   set OPENAI_API_KEY=sk-proj-...
   

   $env:OPENAI_API_KEY="sk-proj-..."
   

   Add this to your shell profile (.bashrc, .zshrc, etc.) to make it permanent.

3. Alternative: Use python-dotenv

   For better security, use python-dotenv to load your API key from a .env file:

   pip install python-dotenv
   

   Create a .env file in your project root:

   .env

   OPENAI_API_KEY=sk-proj-...
   

   Load it in your Python code:

   from dotenv import load_dotenv
   from openai import OpenAI
   
   load_dotenv()  # Load .env file
   client = OpenAI()  # API key automatically loaded from environment
   

   Never commit your .env file to version control. Add it to .gitignore.

4. Verify installation

   Test your installation:

   import openai
   from openai import OpenAI
   
   print(f"OpenAI SDK version: {openai.__version__}")
   
   client = OpenAI()
   print("✓ SDK installed and configured successfully!")
   

Additional Environment Variables

The SDK supports several optional environment variables:

Variable	Description	Default
OPENAI_API_KEY	Your OpenAI API key	Required
OPENAI_BASE_URL	Custom API endpoint	https://api.openai.com/v1
OPENAI_ORG_ID	Organization ID	None
OPENAI_PROJECT_ID	Project ID	None
OPENAI_WEBHOOK_SECRET	Webhook verification secret	None
OPENAI_LOG	Enable logging (info or debug)	Disabled

You can also pass these as parameters when creating the client instead of using environment variables.

Azure OpenAI

To use the SDK with Azure OpenAI, use the AzureOpenAI class:

from openai import AzureOpenAI

client = AzureOpenAI(
    api_version="2023-07-01-preview",
    azure_endpoint="https://example-endpoint.openai.azure.com",
)

completion = client.chat.completions.create(
    model="deployment-name",
    messages=[{"role": "user", "content": "Hello!"}],
)

Azure-specific environment variables:

• AZURE_OPENAI_API_KEY
• AZURE_OPENAI_ENDPOINT
• OPENAI_API_VERSION
• AZURE_OPENAI_AD_TOKEN

Upgrading

To upgrade to the latest version:

pip install --upgrade openai

Check your installed version:

import openai
print(openai.__version__)

Next Steps

Quickstart

Make your first API call with the OpenAI SDK