Realtime `ResponseOutputItemDoneEvent` model has incorrect type for the `item` field

[fedirz]

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

ResponseOutputItemDoneEvent doesn't support item.content[0].type to have a value of "audio" (the value returned by OpenAI's API)

To Reproduce

from openai.types.beta.realtime import ConversationItem, ConversationItemContent, ResponseOutputItemDoneEvent

ResponseOutputItemDoneEvent(
    event_id="event_id",
    response_id="response_id",
    type="response.output_item.done",
    output_index=0,
    item=ConversationItem(content=[ConversationItemContent(type="audio")]),
)

Output:

Traceback (most recent call last):
  File "/home/nixos/code/speaches/src/response_output_item_done_example.py", line 8, in <module>
    item=ConversationItem(content=[ConversationItemContent(type="audio")]),
                                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/nixos/code/speaches/.venv/lib/python3.12/site-packages/pydantic/main.py", line 214, in __init__
    validated_self = self.__pydantic_validator__.validate_python(data, self_instance=self)
                     ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
pydantic_core._pydantic_core.ValidationError: 1 validation error for ConversationItemContent
type
  Input should be 'input_text', 'input_audio', 'item_reference' or 'text' [type=literal_error, input_value='audio', input_type=str]
    For further information visit https://errors.pydantic.dev/2.10/v/literal_error

Code snippets

OS

Linux

Python version

Python 3.12.5

Library version

openai 1.60.0