Show HN: Python lib to run evals across providers: OpenAI, Anthropic, etc.

Library makes requests asynchronously across models, so you can spend a lot of $$ quickly if you want XD. But seriously I hope this enables folks to create and run evals (especially safety ones) a lot easier than before.