Apple supplier Foxconn unveils its first AI model for use in manufacturing

Foxconn, one of the world’s largest contract electronics manufacturers, has launched its inaugural large language model (LLM), “FoxBrain,” marking a significant step in the company's integration of artificial intelligence (AI) into its core operations.
As per a report by news agency Reuters, the company plans to leverage FoxBrain to enhance efficiency in manufacturing and streamline supply chain management.

FoxBrain was developed using 120 of Nvidia's H100 GPUs, completing the training process in approximately four weeks, the company said, adding that the model is built upon Facebook-parent Meta’s Llama 3.1 architecture.
Notably, FoxBrain is the first LLM in Taiwan specifically optimised for traditional Chinese and Taiwanese language styles, incorporating reasoning capabilities, the report said. Foxconn's internal testing indicates that while there is a minor performance difference compared to China's DeepSeek distillation model, FoxBrain's overall performance closely aligns with leading global standards, the company reportedly said.
Initially, FoxBrain will be deployed for internal applications, encompassing data analysis, decision support, document collaboration, mathematical reasoning, problem-solving, and code generation. F
Foxconn aims to expand the model's applications through collaborations with technology partners, sharing open-source information, and promoting AI adoption in manufacturing, supply chain management, and intelligent decision-making processes.
Nvidia played a crucial role in the development of FoxBrain, providing support through its Taiwan-based supercomputer, “Taipei-1,” and offering technical consulting during the model's training phase. Taipei-1, the largest supercomputer in Taiwan, is owned and operated by Nvidia in Kaohsiung.
Foxconn intends to provide further details about FoxBrain during Nvidia's upcoming GTC developer conference in mid-March.