Shipping containers full of B300s + open weight models + Palantir AIP running anywhere in the world managed by Apollo 🔥 We showed this off last fall as part of our Sovereign AI OS reference architecture. Everyone is talking, we have been delivering.

I have seen a few posts about Palantir supporting Chinese open source models, so let me be very clear: this is NOT what we are saying. We are saying that we should and are investing in U.S. based open source models; not only by making open-weight models like Nemotron from NVIDIA accessible across our platforms, but also by providing the tools to evaluate, tune prompts, change architecture, and run all combinations across model providers and families. The tool is called Evolve. This also extends to what we just released with NVIDIA as part of our Sovereign AI OS, called the Intelligent Engine for Open Weight models: This Intelligent Engine allows customers to run Nemotron models in sovereign environments with our critical infrastructure products….AIP, Ontology, Foundry, and Apollo. It will also enable customers to fine-tune these open-weight models for their domain-specific workflows, all securely inside their controlled perimeter. Combined with our domain-optimized harnesses around the Ontology, Nemotron models deliver frontier capabilities for specialized workloads while enabling organizations to retain control of their data, intellectual property, and AI systems.

The AI era will not be won by the model with the best demo. It will be won by the organizations that can operationalize AI securely, govern it in real time, and make decisions faster than the threat landscape evolves. As cyber offense becomes autonomous, defense has to become operational. That’s the shift happening now. This is why what we’re building at @PalantirTech matters.