[Bug] Custom Azure OpenAI deployments doesn't works with `azure.ai.inference` - ResourceNotFoundError (404)

[mahimairaja]

• Package Name: azure.ai.inference
• Package Version: 1.0.0b9
• Operating System: mac OS Sonama 14.5
• Python Version: 3.13, 3.12, 3.11, 3.10

Describe the bug
I am able to access the chat completion api from AzureOpenAI but not using azure.ai.inference

To Reproduce
Steps to reproduce the behavior:

1. create a virtual env
2. install the package using pip install azure-ai-inference

With from openai import AzureOpenAI works!

# sample code to connect to azure openai

import os
from openai import AzureOpenAI

API_KEY = os.getenv("AZURE_OPENAI_API_KEY")
API_VERSION = os.getenv("AZURE_OPENAI_API_VERSION")
AZURE_ENDPOINT = os.getenv("AZURE_OPENAI_ENDPOINT")
MODEL_NAME = os.getenv("AZURE_OPENAI_MODEL_NAME")

client = AzureOpenAI(
    api_key=API_KEY,
    api_version=API_VERSION,
    azure_endpoint=AZURE_ENDPOINT,
)

response = client.chat.completions.create(
    model=MODEL_NAME,
    messages=[{"role": "user", "content": "Hello, world!"}],
)

print(response)

With from azure.ai.inference import ChatCompletionsClient does not work

import os
from azure.ai.inference import ChatCompletionsClient
from azure.core.credentials import AzureKeyCredential

API_KEY = os.getenv("AZURE_OPENAI_API_KEY")
API_VERSION = os.getenv("AZURE_OPENAI_API_VERSION")
AZURE_ENDPOINT = os.getenv("AZURE_OPENAI_ENDPOINT")
MODEL_NAME = os.getenv("AZURE_OPENAI_MODEL_NAME")

client = ChatCompletionsClient(
    endpoint=AZURE_ENDPOINT,
    credential=AzureKeyCredential(API_KEY),
    api_version=API_VERSION,
)

model_info = client.get_model_info()

print(model_info)

Expected behavior
Both needs to work as expected

Error Message

Traceback (most recent call last):
  File "test.py", line 19, in <module>
    model_info = client.get_model_info()
                 ^^^^^^^^^^^^^^^^^^^^^^^
  File ".venv/lib/python3.12/site-packages/azure/core/tracing/decorator.py", line 119, in wrapper_use_tracer
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File ".venv/lib/python3.12/site-packages/azure/ai/inference/_patch.py", line 762, in get_model_info
    raise error
  File ".venv/lib/python3.12/site-packages/azure/ai/inference/_patch.py", line 759, in get_model_info
    self._model_info = self._get_model_info(**kwargs)  # pylint: disable=attribute-defined-outside-init
                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File ".venv/lib/python3.12/site-packages/azure/core/tracing/decorator.py", line 119, in wrapper_use_tracer
    return func(*args, **kwargs)
           ^^^^^^^^^^^^^^^^^^^^^
  File ".venv/lib/python3.12/site-packages/azure/ai/inference/_operations/_operations.py", line 472, in _get_model_info
    map_error(status_code=response.status_code, response=response, error_map=error_map)
  File ".venv/lib/python3.12/site-packages/azure/core/exceptions.py", line 163, in map_error
    raise error
azure.core.exceptions.ResourceNotFoundError: (404) Resource not found
Code: 404
Message: Resource not found

Additional context

[github-actions]

Thanks for the feedback! We are routing this to the appropriate team for follow-up. cc @trrwilson.

[github-actions]

Hello @mahimairaja. I'm an AI assistant for the azure-sdk-for-python repository. I have some suggestions that you can try out while the team gets back to you.

• The behavior you’re seeing is expected when using an Azure OpenAI endpoint with the ChatCompletionsClient.get_model_info method. This method is designed for Serverless API or Managed Compute endpoints and is not supported for Azure OpenAI (or GitHub Models) endpoints.

• For Azure OpenAI deployments, we recommend using the OpenAI client (from openai import AzureOpenAI), as shown in your working sample, which is currently the supported approach.

• For more details see: azure-ai-inference README (refer to the section on model information limitations) and Azure OpenAI API Specs for version details.

The team will get back to you shortly, hopefully this helps in the meantime.

[github-actions]

Thanks for the feedback! We are routing this to the appropriate team for follow-up. cc @dargilco @jhakulin @trangevi.