We are grateful to Prime Intellect for their generous support in running proprietary model APIs through their Inference platform.