Palantir, Nvidia bring Nemotron open models to sovereign AI

Palantir said it will run Nvidia's Nemotron models across its platforms in sovereign environments in US government agencies and critical infrastructure.

The move is notable given enterprises are facing stick shock over AI and token costs and looking for open models to trim expenses. Many of these organizations don't want to (or are restricted) from using Chinese open models.

In the US, Nvidia's Nemotron models are emerging as one of the top choices for open models.

• Why enterprise AI leaders need to bank on open-source LLMs
• Nvidia launches Nemotron 3 open models to enable multi-agent systems
• Nvidia's model parade: Llama Nemotron, Cosmos additions, Isaac GROOT N1
• Nvidia Nemotron: Much needed open-source model champion in US

Under an expanded partnership, Palantir will create an engine that runs Nvidia's AI stack and Nemotron models in sovereign deployments that use Palantir's AIP, Ontology, Foundry and Apollo platforms. Palantir will run the Nemotron models in its optimized harnesses focused on industries.

According to Palantir, the Nvidia Nemotron options enable customers to control their data, IP and AI systems with data authorization, rights to erasure and auditability.

Palantir and Nvidia will offer deployment, context and model engineering. Palantir and Nvidia previously announced a sovereign AI operating system reference architecture.

• Palantir and Nvidia build sovereign, on-premises AI reference architecture

Separately Palantir said it expanded its partnership with Surf Air Mobility, an air mobility platform. Surf Air Mobility's SurfOS is built on Palantir's AIP and Foundry.