AuthenticationError: Incorrect API key provided: https://************************************************.com.

[hherpa]

I'm generating an openai api format link for wizardcoder-13b to use in autogen with support for downloading various documents, but your google collab notebook doesn't accept it.

---------------------------------------------------------------------------
AuthenticationError                       Traceback (most recent call last)
[<ipython-input-24-85f187986a5e>](https://localhost:8080/#) in <cell line: 5>()
      3     embedding_function=OpenAIEmbeddings()
      4 )
----> 5 vectorstore.add_documents(docs)

16 frames
[/usr/local/lib/python3.10/dist-packages/openai/api_requestor.py](https://localhost:8080/#) in _interpret_response_line(self, rbody, rcode, rheaders, stream)
    773         stream_error = stream and "error" in resp.data
    774         if stream_error or not 200 <= rcode < 300:
--> 775             raise self.handle_error_response(
    776                 rbody, rcode, resp.data, rheaders, stream_error=stream_error
    777             )

AuthenticationError: Incorrect API key provided: https://************************************************.com. You can find your API key at https://platform.openai.com/account/api-keys.

Is there anything I can do about it? (+ if you know another way to generate OpenAI API format link, I would be very grateful if you could share it). Also if the error isn't in the link, I can assume it's in the way I'm uploading it (I edited the code a bit):

import autogen
import os

config_list = [
    {
        "model": "mistral-7b",
        "api_key": "https://brothers-chip-training-potentially.trycloudflare.com", # - This is roughly what the links I generate look like (regular autogen accepts them and works fine, but I need document support)
    },
]

gpt4_api_key = "https://brothers-chip-training-potentially.trycloudflare.com"
os.environ['OPENAI_API_KEY'] = gpt4_api_key

Once again, I would like to point out that normal autogen accepts and works with this link, but with LangChain it does not allow it for some reason.