Cannot use verbosity for GPT-5 API

[hgnzheng]

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

When we use OpenAI version 1.70.0, we are getting the following error:

TypeError: Responses.create() got an unexpected keyword argument 'verbosity'

We we used the most up-to-date OpenAI version of 1.99.3, we get the following:

BadRequestError: Error code: 400 - {'error': {'message': "Unknown parameter: 'verbosity'.", 'type': 'invalid_request_error', 'param': 'verbosity', 'code': 'unknown_parameter'}}

Is there a fix to this?

To Reproduce

1. Download either OpenAI package with version 1.70.0 or 1.99.3.
   2.Run the following code.

Code snippets

response = client.responses.create(
    model="gpt-5-2025-08-07",
    input=[
        {"role": "user", "content": 'Hello'},
    ],
    reasoning={
        "effort": "minimal",
    },
    max_output_tokens=4096,
    stream=False,
    verbosity='low'
)

OS

macOS

Python version

Python v3.10.14

Library version

openai v1.70.0

[hgnzheng]

Fixed.

[inesruizruiz]

I also get this error with gpt-5-mini, I am using the latest version released==1.99.6.
TypeError: Responses.create() got an unexpected keyword argument 'verbosity'

Code snippet:
response = client.responses.create( model="gpt-5-mini", instructions=prompt, input=messages, reasoning={"effort": "low"}, verbosity="low" )

[RobertCraigie]

you have to pass it to the text param, not at the top level, e.g.

            text={
                "format": {"type": "text"},
                "verbosity": "low",
            },

[hayescode]

you have to pass it to the text param, not at the top level, e.g.

        text={
            "format": {"type": "text"},
            "verbosity": "low",
        },

@RobertCraigie with this snippet it's valid and seems to work, but no matter what value we set for verbosity the Response object is always null.

from openai.types.responses import ResponseTextConfigParam, ResponseFormatTextConfigParam
from openai.types.responses.response_format_text_config_param import ResponseFormatText

stream = await llm.responses.create(
    model="gpt-5",
    input="what is the answer to the universe?",
    stream=True,
    text=ResponseTextConfigParam(
        format=ResponseFormatText(type="text"),
        verbosity="low"
    )
)

async for event in stream:
    if event.type == 'response.created':
        print(event.response.text)
    print(event.model_dump_json(indent=2))

ResponseTextConfig(format=ResponseFormatText(type='text'), verbosity=None)
    "text": {
      "format": {
        "type": "text"
      },
      "verbosity": null
    },

[RobertCraigie]

And it isn't anywhere else in the response object? That sounds like an API issue so would you mind reposting at community.openai.com?