The aim is to develop an open, sovereign European ensemble of large language and reasoning models (LLMs/LRMs) and agent-based systems that are powerful, transparent, and deployable on European cloud and edge infrastructures.GitHub: https://github.com/soofi-project