Azure AI Inference SDK fails to complete chat completion with gpt-4.1

[charendt]

• Package Name: azure-ai-inference
• Package Version: 1.0.0b9
• Operating System: Windows 11
• Python Version: 3.12.9

Describe the bug
Chat completion request fails with

ERROR:
JSONDecodeError: Expecting value: line 1 column 1 (char 0)

To Reproduce
Steps to reproduce the behavior:

client = ChatCompletionsClient(
    endpoint=endpoint,
    credential=AzureKeyCredential(os.getenv("AZURE_OPENAI_API_KEY"))    
)

response = client.complete(
    messages=[
        SystemMessage(content="You are a helpful assistant."),
        UserMessage(content="Where is Zurich?")
    ],
    model="gpt-4.1"
)

Expected behavior
Return the chat completion result

Screenshots
n/a

Additional context
This works fine with any other AOAI model, it only fails with gpt-4.1

[kashifkhan]

Thank you for the feedback @charendt. Ill route this over to the right team who can assist you better,

[github-actions]

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

[trangevi]

@charendt thanks for the report. I'm able to repro the error and am investigating further.

In the meantime, you may be able to work around the issue by using openai.azure.com endpoint that is associated with your model deployment. It should look something like https://<servicename>.openai.azure.com/openai/deployments/gpt-4.1, and you should be able to see it as the "Target URI" for your model in the Foundry UI. You'll just need to remove everything after "gpt-4.1" in that URI, and then you can also remove the model parameter to the complete call.

[ytuvia]

I have the same issue, workaround works. waiting for your correction