Nvidia and Meta have committed to deepening their AI partnership, with a focus on improving Muse agent performance and agentic security and backed by a $3 billion Nvidia investment.

Under new plans, Meta will use Nvidia’s Vera Rubin data centre platform, including its Rubin GPUs and Vera CPUs, for future AI agents.

Meta will also deploy Nvidia’s ethernet networking platform, used for the massive data throughput required for agentic AI inference, alongside its data processing unit BlueField-4.

Meta’s Muse personal AI agents, which autonomously complete tasks for users across a range of apps, leverage Muse AI models trained on Nvidia hardware and use Nvidia infrastructure for processing.

Nvidia will invest $3 billion directly into Meta’s equity offering.

“AI agents represent a new computing paradigm – software that understands our goals, reasons, uses tools and gets work done on our behalf,” said Jensen Huang, founder and chief executive of Nvidia.

“This changes how we use computers, from operating applications to working with intelligent agents. Nvidia and Meta are expanding our partnership to optimize the computing infrastructure end to end for capable, efficient and secure AI agents, so Muse can empower billions of people to achieve more and pursue greater ambitions.”

The news builds on the multiyear chip deal Meta signed with Nvidia in February, part of a massive infrastructure push by the social media giant to build a lead in the AI market.

Fruits of its data centre investment to date include Muse Wonder and Wonder Full. The company claims these new flagship models outperform competing offerings such as Anthropic’s Claude Opus 5.5, OpenAI’s GPT-6 Astra and Google’s Gemini 4 Argon .

Outside of their formal partnership, Nvidia and Meta have aligned in their rejection of calls by figures such as Dario Amodei to slow down AI development.

---