Nvidia Corp (NASDAQ:NVDA, ETR:NVD) announced it is teaming up with Perplexity to roll out across Europe locally optimized large language models (LLMs) tailored to the continent's diverse languages and industries.
Unveiled at Nvidia’s GTC event in Paris, the collaboration involves a wide coalition of European and Middle Eastern research institutions, AI developers, and regional cloud providers.
The partners will use Nvidia's Nemotron model-building framework and DGX Cloud Lepton infrastructure to create and distribute LLMs that support Europe's 24 official languages and reflect distinct cultural contexts.
Nvidia CEO Jensen Huang highlighted the strategic importance of this effort in achieving digital sovereignty, particularly for countries with limited existing AI infrastructure or underrepresented languages.
He noted that Nvidia is focused on creating “reasoning models” by generating high-quality synthetic data to boost performance in languages with limited training corpora.
“Perplexity will help distribute these models across Europe,” Huang said, noting that Germany is already “a significant market” for the technology.
Huang also personally endorsed Perplexity, stating that he uses the platform “almost daily” for research purposes.
“Perplexity’s goal is to provide accurate, trustworthy answers to any question from any person, wherever they are,” Perplexity CEO and cofounder Aravind Srinivas said in a statement.
“Bringing NVIDIA-optimized sovereign AI models to Perplexity empowers innovation in Europe with AI built and running in the region.”
Nvidia shares were little changed on the announcement, trading hands at $144 on Wednesday morning.