27 ms·

Fair enough. With that said, support for running a local, self-hosted, and non-OpenAI cloud-hosted models is in the works.