How to intermediate · 3 min read

Azure OpenAI function calling

Q: Azure OpenAI function calling

Use the AzureOpenAI client from the openai package with your Azure endpoint and deployment name. Pass functions and function_call parameters in chat.completions.create() to enable function calling in Azure OpenAI.

Quick answer

Use the AzureOpenAI client from the openai package with your Azure endpoint and deployment name. Pass functions and function_call parameters in chat.completions.create() to enable function calling in Azure OpenAI.

PREREQUISITES

Python 3.8+
Azure OpenAI API key
Azure OpenAI endpoint URL
pip install openai>=1.0

Setup

Install the official openai Python package version 1.0 or higher. Set environment variables for your Azure OpenAI API key, endpoint, and deployment name. Use the AzureOpenAI client to connect.

bash

pip install openai>=1.0

Step by step

This example demonstrates how to call a function using Azure OpenAI's function calling feature. Define your function schema, then pass it in the functions parameter. Use function_call="auto" to let the model decide when to call the function.

python

import os
from openai import AzureOpenAI

client = AzureOpenAI(
    api_key=os.environ["AZURE_OPENAI_API_KEY"],
    azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
    api_version="2024-02-01"
)

def main():
    functions = [
        {
            "name": "get_current_weather",
            "description": "Get the current weather in a given location",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {"type": "string", "description": "The city and state, e.g. San Francisco, CA"},
                    "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}
                },
                "required": ["location"]
            }
        }
    ]

    messages = [
        {"role": "user", "content": "What is the weather like in Seattle?"}
    ]

    response = client.chat.completions.create(
        model=os.environ["AZURE_OPENAI_DEPLOYMENT"],
        messages=messages,
        functions=functions,
        function_call="auto"
    )

    message = response.choices[0].message
    print("Model response:", message.content)
    if message.function_call:
        print("Function call requested:", message.function_call)

if __name__ == "__main__":
    main()

output

Model response: 
Function call requested: {'name': 'get_current_weather', 'arguments': '{"location": "Seattle", "unit": "fahrenheit"}'}

Common variations

Use function_call="none" to disable function calling.
Specify function_call={"name": "function_name"} to force a specific function call.
Use async Python with asyncio and await on client.chat.completions.acreate().
Change the model deployment name via environment variable for different Azure OpenAI models.

python

import asyncio
import os
from openai import AzureOpenAI

client = AzureOpenAI(
    api_key=os.environ["AZURE_OPENAI_API_KEY"],
    azure_endpoint=os.environ["AZURE_OPENAI_ENDPOINT"],
    api_version="2024-02-01"
)

async def async_main():
    functions = [
        {
            "name": "get_current_weather",
            "description": "Get the current weather in a given location",
            "parameters": {
                "type": "object",
                "properties": {
                    "location": {"type": "string"},
                    "unit": {"type": "string", "enum": ["celsius", "fahrenheit"]}
                },
                "required": ["location"]
            }
        }
    ]

    messages = [{"role": "user", "content": "Tell me the weather in Boston."}]

    response = await client.chat.completions.acreate(
        model=os.environ["AZURE_OPENAI_DEPLOYMENT"],
        messages=messages,
        functions=functions,
        function_call={"name": "get_current_weather"}
    )

    message = response.choices[0].message
    print("Async model response:", message.content)
    if message.function_call:
        print("Function call:", message.function_call)

if __name__ == "__main__":
    asyncio.run(async_main())

output

Async model response: 
Function call: {'name': 'get_current_weather', 'arguments': '{"location": "Boston", "unit": "celsius"}'}

Troubleshooting

If you get authentication errors, verify your AZURE_OPENAI_API_KEY and AZURE_OPENAI_ENDPOINT environment variables.
If function calling is ignored, ensure your model deployment supports function calling and you use api_version="2024-02-01" or later.
Check that your function schema is valid JSON Schema and includes required fields.

✅

Key Takeaways

Use AzureOpenAI client with functions and function_call params to enable function calling.
Set api_version="2024-02-01" or later for function calling support in Azure OpenAI.
You can control function calls with function_call="auto", "none", or specify a function name.
Async calls are supported via acreate() method on the client.
Validate your function JSON schema carefully to avoid errors.

Verified 2026-04 · gpt-4o, AzureOpenAI

Verify ↗