ChatCompletionStreamManager object does not support the asynchronous context manager protocol

[lucashofer]

Confirm this is an issue with the Python library and not an underlying OpenAI API

• [X] This is an issue with the Python library

Describe the bug

The docs here say that the following should be possible

import openai
import asyncio

async def test_streaming():
    client = openai.OpenAI()

    async with client.beta.chat.completions.stream(
        model='gpt-4o-2024-08-06',
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": "Tell me a joke."},
        ],
    ) as stream:
        async for event in stream:
            if event.type == 'content.delta':
                print(event.delta, flush=True, end='')
            elif event.type == 'content.done':
                print("\nContent generation complete.")
                break

# Run the streaming test
asyncio.run(test_streaming())

However, this gives

TypeError: 'ChatCompletionStreamManager' object does not support the asynchronous context manager protocol

When I run without async it works fine ie

import openai

def test_streaming():
    client = openai.OpenAI()

    with client.beta.chat.completions.stream(
        model='gpt-4o-2024-08-06',
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": "Tell me a joke."},
        ],
    ) as stream:
        for event in stream:
            if event.type == 'content.delta':
                print(event.delta, flush=True, end='')
            elif event.type == 'content.done':
                print("\nContent generation complete.")
                break

# Run the streaming test
test_streaming()

To Reproduce

Run the above code snippet which is the beta async chat_completion (and should handle the new pydantic parsing)

Code snippets

OS

macOS

Python version

Python 3.11-3.12

Library version

1.40.4

[RobertCraigie]

ah @lucashofer, sorry those docs don't make it clear, you have to use AsyncOpenAI() for async requests.

for example

import openai
import asyncio

async def test_streaming():
    client = openai.AsyncOpenAI()

    async with client.beta.chat.completions.stream(
        model='gpt-4o-2024-08-06',
        messages=[
            {"role": "system", "content": "You are a helpful assistant."},
            {"role": "user", "content": "Tell me a joke."},
        ],
    ) as stream:
        async for event in stream:
            if event.type == 'content.delta':
                print(event.delta, flush=True, end='')
            elif event.type == 'content.done':
                print("\nContent generation complete.")
                break

# Run the streaming test
asyncio.run(test_streaming())