MAX Engine

MAX is a high-performance inference engine developed by Modular, designed to optimize the deployment of AI models across various hardware architectures. The brand was founded to address the fragmentation in AI infrastructure, providing a unified platform that bridges the gap between heterogeneous hardware and complex software frameworks.

Originally created by Modular AI, Inc. in the United States, the brand remains under its primary ownership. The engine is engineered and maintained primarily in the United States, though its community-driven components and development teams operate globally. It is recognized for its ability to enhance model latency and throughput, particularly through its use of Mojo and advanced compiler technologies.