Tags: #inference-engine

MNN is a blazing-fast, lightweight deep learning inference engine optimized for high-performance on-device AI and Large Language Models.

A universal machine learning compiler and high-performance deployment engine for large language models, enabling native execution across diverse platforms.