06:32 AM EDT, 03/10/2025 (MT Newswires) -- Apple's ( AAPL ) supplier Foxconn, also known as Hon Hai Precision Industry, said Monday it launched its first large language model, 'FoxBrain,' designed to improve manufacturing and supply chain management.

The company said the model, optimized for traditional Chinese and Taiwanese language styles, was trained on 120 Nvidia ( NVDA ) H100 GPUs and completed in about four weeks.

Based on Meta's (META) Llama 3.1 architecture, FoxBrain was initially designed for internal applications, including data analysis, decision support, document collaboration, and code generation, the company added.

While it still slightly trails DeepSeek's distillation model, Foxconn said its performance is nearly on par with leading global standards.

Foxconn said it plans to collaborate with technology partners to expand FoxBrain's applications and promote AI in manufacturing and intelligent decision-making.

Nvidia ( NVDA ) supported the model's development through its Taipei-1 Supercomputer and technical consultation during model training, it added.