Nvidia said on October 26, 2026 that it will invest $3 billion in Meta’s equity offering and expand its collaboration with Meta on Muse, Meta’s family of AI models and agents. The announcement came the same day Meta disclosed plans to raise $95 billion in new stock to pay for AI infrastructure, and it ties the world’s largest AI chip supplier more closely to one of its biggest customers.

What Nvidia and Meta announced

According to Nvidia’s release, the companies will work together to improve inference and training performance for Muse and the Muse family of models, citing demand for useful AI agents. They plan to optimize across workflows to raise performance and lower total cost of ownership. Nvidia says Meta’s Muse models and agents, which run on Nvidia infrastructure, have delivered “frontier-level benchmark performance” and rising consumer downloads since launch; those are the companies’ own characterizations and the release cites no figures.

Meta intends to expand its deployment of Nvidia infrastructure, including Vera Rubin GPUs and Vera CPUs for what it calls next-generation agents, with a focus on performance per watt. It will also deploy Nvidia Spectrum-X Ethernet networking and BlueField-4 data processing units to improve network efficiency and throughput.

The $3 billion investment

The headline number is Nvidia’s $3 billion stake in Meta’s equity offering. Our earlier report on the $95 billion stock offering noted that The Guardian had named Nvidia, AMD and Broadcom as potential cornerstone investors who could together buy up to $5 billion of the shares. Nvidia’s own release now puts its commitment at $3 billion, which is within the $5 billion that The Guardian said the three chipmakers could buy together. The release carries the usual forward-looking caveats about completion and closing conditions, so the investment is an intention until the offering closes.

Exterior of Nvidia's headquarters building in Santa Clara, California
Nvidia, whose headquarters is pictured, will invest in Meta's offering while supplying the chips Meta plans to buy. Photo: Coolcaesar / Wikimedia Commons, CC BY-SA 4.0

The circularity is worth stating plainly. A chip supplier is buying shares in a customer that will use the proceeds, in part, to buy more of that supplier’s chips. That pattern is not new in AI, but it makes the arrangement a financing story as much as a technology one. Meta’s offering is meant to cover infrastructure spending, and the Nvidia stake helps make that raise easier to place.

Agent security tools on the table

Beyond hardware, the two companies said they will consider further collaboration on Nvidia’s OpenShell Reference Architecture. That includes OpenShell for safe deployment of agents and Nvidia Sentry for monitoring agent behavior, plus data-processing work across structured, unstructured and graph data. The wording is exploratory: the companies “will consider” these options, so no deployment is committed.

Nvidia chief executive Jensen Huang said agents represent a new computing paradigm, while Meta chief executive Mark Zuckerberg described Muse as the way Meta brings “personal superintelligence” to everyone. For background on the model family, see our report on Muse Spark’s open weights.

Why it matters

Three things stand out. Meta is committing to Nvidia’s newest CPU and GPU generations rather than relying only on in-house silicon. Nvidia is using its balance sheet to support demand for its products. And the focus on agents, rather than chatbots, shows where both firms expect inference spending to go. Readers should note that none of the performance or cost claims in the release have been independently measured.

What to watch next

  • Pricing and closing of Meta’s $35 billion underwritten offering and the size of Nvidia’s final allocation.
  • Whether Meta publishes Muse benchmarks or download figures to back the “frontier-level” claim.
  • Any deployment details for OpenShell and Sentry inside Meta.
  • Reaction from AMD and Broadcom, also named as possible cornerstone investors.
  • Regulatory or investor scrutiny of supplier-customer equity ties.

Key takeaways

  • Nvidia will invest $3 billion in Meta’s equity offering, according to Nvidia.
  • The companies are expanding work on Muse models, agents and inference efficiency.
  • Meta plans to deploy more Vera Rubin GPUs, Vera CPUs, Spectrum-X and BlueField-4.
  • OpenShell and Sentry agent-security tools are under consideration, not committed.
  • Performance claims for Muse come from the companies and are not independently verified.

Frequently asked questions

How much is Nvidia investing in Meta?

Nvidia says it will invest $3 billion in Meta’s equity offering. The release notes the investment is subject to completion and closing conditions.

What is Muse?

Muse is Meta’s family of AI models and agents, which Meta presents as its route to personal superintelligence. Nvidia says Muse runs on its infrastructure.

Why would a chipmaker buy shares in its customer?

It supports the customer’s financing for AI infrastructure, which in turn supports demand for the chipmaker’s products. The structure ties the two companies’ fortunes more closely together.