1.16.0 breaks response_format

[arthurbrenno]

I'm using the latest version of the cerebras-cloud-sdk. Using the new "ResponseFormatResponseFormatJsonSchemaJsonSchemaTyped", it gives me the following error:

BadRequestError: Error code: 400 - {'message': 'Unsupported "response_format" type: "json_schema"', 'type': 'invalid_request_error', 'param': 'response_format', 'code': 
'wrong_api_format'}

relevant part of my code:

await client.chat.completions.create(
                messages=[message.to_cerebras_format() for message in messages],
                model=self.model_name,
                max_completion_tokens=max_completion_tokens,
                n=n,
                stop=list(stop_sequences) if stop_sequences else None,
                response_format=ResponseFormatResponseFormatJsonSchemaTyped(
                    json_schema=ResponseFormatResponseFormatJsonSchemaJsonSchemaTyped(
                        name=response_model.__name__,
                        description="Structured Response",
                        schema=flatten_msgspec_schema(
                            msgspec.json.schema(response_model), openai_like=True
                        ),
                        strict=True,
                    ),
                    type="json_schema",
                )
                if response_model
                else None,
                temperature=temperature,
                tools=[
                    ToolTyped(
                        function=Function.from_callable(tool).to_cerebras_function(),
                        type="tool",
                    )
                    for tool in tools
                ]
                if tools
                else None,
                parallel_tool_calls=True,
            ),
        )

Am I doing something wrong? I see that I have no docs to look in. So I'm kinda going by my own and see where I can get.