(401) The Creates a model response for the given chat conversation. Operation under Azure AI Model Inference API version 2024-05-01-preview is not supported with the current subscription key and pricing tier OpenAI.S0.

Question

Hi, I'm trying to implement a Python app with llms served from Azure AI Foundry. I have the following code:

from azure.ai.projects import AIProjectClient
from azure.identity import DefaultAzureCredential

project_connection_string = "<project-connection-str>"

project = AIProjectClient.from_connection_string(
    conn_str=project_connection_string, credential=DefaultAzureCredential()
)

chat = project.inference.get_chat_completions_client(connection_name="<connection-name>")

response = chat.complete(
    model="<model-name>",
    messages=[
        {"role": "system","content": "You are an AI assistant that speaks like a techno punk rocker from 2350. Be cool but not too cool. Ya dig?",},
        {"role": "user", "content": "Hey, can you help me with my taxes? I'm a freelancer."},
    ],
)

print(response.choices[0].message.content)

The error occurs during the chat.complete() method, and I don't know how to fix it. I want to connect through Foundry like this so I can use both open-source or openai models, instead of using AzureOpenAI objects (which works)

The following alternative results in the same error:

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

# Datos de conexión
endpoint = "<endpoint>"  # Cambia por el tuyo
api_key = "<api-key>"  # Reemplaza con tu clave real

# Crear cliente de inferencia directamente
chat = ChatCompletionsClient(endpoint=endpoint, credential=AzureKeyCredential(api_key))

# Hacer la consulta al modelo específico
response = chat.complete(
    model="<model-name>",  # Asegúrate de usar el nombre exacto del deployment
    messages=[{"role": "user", "content": "¿Cómo mejorar mi productividad?"}]
)

# Imprimir la respuesta
print(response.choices[0].message.content)

How can I fix this?