NVIDIA will invest $3 billion in Meta’s equity offering as the two companies deepen their partnership on Muse, Meta’s AI assistant. The joint announcement, made Tuesday, ties the world’s most valuable chipmaker directly to Meta’s push to put an AI agent in front of billions of users, and it lands one day after Meta unveiled a fundraising plan of up to $95 billion.

Under the deal, NVIDIA and Meta will expand their work on Muse and the Muse family of models to improve inference and training performance. Meta will also expand its deployment of NVIDIA’s Vera Rubin platform and Vera CPUs, alongside Spectrum-X Ethernet networking and BlueField-4 data processing units, to run the next generation of AI agents.

What Was Announced

According to the joint announcement, Meta’s Muse models and agents, which run on NVIDIA infrastructure, have delivered frontier-level benchmark performance and surging consumer downloads since launch. The two companies now intend to deepen their collaboration with joint optimizations across workflows, aiming to maximize performance and improve total cost of ownership.

The companies are also considering further work together on NVIDIA’s OpenShell Reference Architecture, including OpenShell for safe deployment and Sentry for monitoring how AI agents behave, as well as structured, unstructured and graph data processing.

“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 CEO 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 $3 Billion Investment

In addition to the technical partnership, NVIDIA will invest $3 billion in Meta’s equity offering. Earlier on Tuesday, Meta increased the size of that offering to $97.5 billion, the largest fundraising in the company’s history, according to Reuters.

The investment puts a concrete number on the chipmaker interest Meta flagged a day earlier. When Meta announced its fundraising plan on Monday, it said NVIDIA, AMD and Broadcom had expressed interest in buying up to $5 billion of the offering in aggregate, on a nonbinding basis. NVIDIA’s $3 billion commitment is the first of those to be confirmed. For the background on Meta’s fundraising plan, see GPTOnline’s report: Meta Plans Up to $95 Billion Stock Sale to Fund AI Buildout, With Nvidia, AMD and Broadcom Circling.

The offering itself consists of concurrent underwritten offerings now, including $22.5 billion of Class A common stock and $12.5 billion of mandatory convertible preferred stock, plus a planned at-the-market program of new non-voting Class C shares expected in the first quarter of 2027. Proceeds are earmarked for AI infrastructure and Meta’s personal superintelligence effort.

The deal at a glance
Graphic: GPTOnline, using NVIDIA and Meta assets

What Muse Is

Muse, known internally at Meta as Hatch, is the company’s AI assistant and the centerpiece of CEO Mark Zuckerberg’s plan to deliver “personal superintelligence” to the billions of people who use Meta’s services each day. It is modeled on the open-source AI agent OpenClaw and can access a user’s email, calendar, payments, health, shopping and smart-home applications, with users able to revoke access at any time.

That breadth of access is exactly why the infrastructure behind Muse matters. An assistant that acts across apps needs fast, cheap inference and strong guardrails, which is what the NVIDIA partnership is built to provide: Vera Rubin hardware for the compute, Spectrum-X networking to move data between systems, and OpenShell and Sentry tooling to deploy agents safely and watch what they do.

What Muse is
Graphic: GPTOnline, using NVIDIA and Meta assets

Why It Matters

The deal cements a two-way dependency at the heart of the AI buildout. Meta needs NVIDIA’s chips and systems to run Muse at the scale of its user base, and NVIDIA needs Meta’s workloads to keep its newest platforms, led by Vera Rubin, in heavy use. A $3 billion equity stake makes that alignment financial as well as technical.

It also sharpens the competitive picture for AI agents. With Anthropic filing to go public the same week and OpenAI cutting prices on its new GPT-6.1 Terra model, the race is no longer only about who builds the smartest model. It is about who can run an agent for billions of people at a cost that works, and who users trust with access to their inbox, calendar and payments. NVIDIA and Meta have now answered that question together, with money on the table.

Sources