How is openai API provided?

in the repo readme, it says “In the prompt engineering track, we provide participants with access to the OpenAI API.” Does it mean we are given an API key?

Another question is what version of openai API is used? I had an error when " from openai import OpenAI", the error is ImportError: cannot import name 'OpenAI' from 'openai' (/raid/data/mambaforge/envs/rapids-23.08-stable/lib/python3.10/site-packages/openai/__init__.py)

The openai version I pip installed is 0.28.0. Thank you!

after reading the readme I think I understand how it works now. basically, we implement the prompt_agent API and a caller will use prompt_agent’s output to call the actual openai api. so participants don’t need API keys.

Is there any solution for our code to directly access the API? Perhaps some of us would like to control how the two calls to api per conversation is used up?

Thank you in advance!