Palantir introduced a new intelligent engine using NVIDIA Nemotron open models to deliver frontier-quality AI to U.S. go

Open source software has long been a pillar of U.S. technology leadership. In 1969, DARPA connected four university computers from UCLA, Stanford, UCSB and the University of Utah, laying the infrastructure backbone that became the internet. In those early days, U.S.-led open source contributions also drove leadership in coding languages, with UNIX in 1969 and C at Bell Labs in 1972. These languages led to more open source software building on those foundations, including the Linux Kernel in 1991, GitHub in 2008 and Docker in 2013.

Today, open models are making frontier-level AI broadly accessible, with control over customization and trust through transparency. They give enterprises and government agencies the ability to inspect, adapt and deploy AI in sensitive environments, making them essential for national security, corporate sustainability and industrial innovation. With domain-optimized harnesses, strong open models can deliver frontier capabilities while helping customers retain control over proprietary data, model weights and deployment environments. Palantir's announcement brings NVIDIA Nemotron open models into air-gapped environments — secure setups that are completely isolated from unsecured networks — on NVIDIA accelerated computing.

Palantir will use NVIDIA Nemotron open models to build custom frontier-quality models to serve the U.S. government. Many of the government's operations mirror private-sector enterprises, including commerce, energy, healthcare, agriculture, education and transportation. With about three million civilian employees, the U.S. government is essentially one of the world's largest enterprises. Providing critical services across so many disciplines is incredibly complex. AI can help streamline this complexity and boost insights to drive productivity.

With this new engine, agencies and operators can run customized Nemotron models on their own infrastructure, train on their own data and retain full ownership of the resulting models, including the weights that encode their operational knowledge. Palantir's Sovereign AI Operating System, built on AIP, Ontology, Foundry and Apollo, handles the operational and data authorization layer for easy deployment in sensitive environments. Explicit data authorization, architecturally enforced isolation and full auditability are already central to the system. As customized models are used in production, agencies and operators can continually improve them within their own environments using new data and feedback, creating a data flywheel that optimizes performance while keeping data, models and auditability under customer control.

The combination delivers three core benefits. Trust comes through transparency, since researchers can independently review open models to identify vulnerabilities, biases and unintended behaviors. Customization and control allow companies, governments and developers to modify and fine-tune open models to suit their use cases and deploy them in regulated environments where closed models might breach data security or privacy laws. Lower costs fuel economic development, with about two-thirds of companies already using open models and reporting cost efficiency. Running the models on air-gapped NVIDIA-powered infrastructure keeps data and models secure while supporting critical government missions.