Are you able to use `strict=True` when using pydantic models? It doesn't seem to be valid for me. I think that only works for json schemas.
class FooBar(BaseModel):
foo: list[str]
bar: list[int]
prompt = """#Task
Your job is to reply with Foo Bar, a json object with foo, a list of strings, and bar, a list of ints
"""
response = openai_client.chat.completions.parse(
model="gpt-5-nano-2025-08-07",
messages=[{"role": "system", "content": FooBar}],
max_completion_tokens=4096,
seed=123,
response_format=CommentAnalysis,
strict=True
)
> TypeError: Completions.parse() got an unexpected keyword argument 'strict'
