Distributed training platform — train, fine-tune and deploy frontier open models across a global network of GPUs, with the INTELLECT model family.