Preferred Networks develops AI technology across four integrated layers: proprietary AI chips (MN-Core), computing infrastructure, generative AI foundation models (PLaMo), and industry-specific AI products and solutions.
AI-dedicated processor series co-developed with Kobe University for deep learning training and inference; MN-Core 2 sold as the MN-Server 2 rack server and MN-Core 2 Devkit workstation.
Family of generative AI foundation models built from scratch, including flagship PLaMo Prime, edge-optimized PLaMo Lite, PLaMo Translate for Japanese text, and finance-tuned PLaMo-fin-prime.
Cloud service built and operated by PFN for deep learning and AI workloads, providing exclusive access to clusters of MN-Core 2 accelerators and a multi-tenant Kubernetes environment.
AI-powered atomistic simulator for materials discovery using the proprietary PFP model; used by over 150 organizations globally including automotive, chemical, and electronics manufacturers.
Autonomous mobile robot developed and marketed by Preferred Robotics, a PFN group company; navigates and transports items in factories, warehouses, offices, and homes. Available in Japan.
AI-based visual inspection software for industrial parts, food, and other manufactured goods, trained using around 100 photographs per part; used by over 200 clients. Available in Japan only.
AI-driven solution trained on plant operation data and expert knowledge for automated, optimized control of process manufacturing plants under complex or disturbed conditions.
Enterprise AI platform designed to help organizations adopt AI across the business, part of PFN's AI Products and Solutions portfolio.