New responses API does not work with AzureOpenAI #2184
Description
Confirm this is an issue with the Python library and not an underlying OpenAI API
- This is an issue with the Python library
Describe the bug
On the latest version of openai==1.66.2, I have this current code that works well:
import os
from openai import AzureOpenAI
client = AzureOpenAI(
api_version=os.getenv("OPENAI_API_VERSION"),
api_key=os.getenv("OPENAI_API_KEY"),
azure_endpoint=os.getenv("OPENAI_AZURE_ENDPOINT"),
)
completions = client.chat.completions.create(
model=os.getenv("OPENAI_AZURE_DEPLOYMENT"),
messages=[
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "What is the capital of the United States?"},
],
)
print(completions.choices[0].message.content)
# returns 'The capital of the United States is Washington, D.C.'I wanted to test the new responses API, so I followed the docs:
import os
from openai import AzureOpenAI
client = AzureOpenAI(
api_version=os.getenv("OPENAI_API_VERSION"),
api_key=os.getenv("OPENAI_API_KEY"),
azure_endpoint=os.getenv("OPENAI_AZURE_ENDPOINT"),
)
response = client.responses.create(
model=os.getenv("OPENAI_AZURE_DEPLOYMENT"),
instructions="You are a coding assistant that talks like a pirate.",
input="How do I check if a Python object is an instance of a class?",
)
print(response.output_text)But this results in:
openai.NotFoundError: Error code: 404 - {'error': {'code': '404', 'message': 'Resource not found'}}
Am I missing something or is this just not supported yet? Note that the deployed model on Azure is GPT-4o.
To Reproduce
All information required is above.
Code snippets
OS
Fedora
Python version
Python3.10
Library version
openai==1.66.2