Data sourced from Dakota Private Markets, the private fund performance platform powered by Dakota. Learn More | Request Access
Nvidia will invest $3 billion in Meta's equity offering, part of a plan by Meta to raise up to $95 billion for AI infrastructure. The investment comes with an expanded partnership on Muse, Meta's family of AI models and its personal AI agent, aimed at improving training and inference performance.
The announcement on October 27, 2026 turns one of Meta's largest suppliers into one of its shareholders. Meta also said Nvidia, AMD and Broadcom expressed interest in buying up to $5 billion of the shares combined, which means three chip suppliers may end up owning a piece of the company buying their products.
The deal sits inside a much larger financing. Meta expects 2026 capital expenditures of $135 billion to $145 billion, and the equity raise is how it intends to fund part of that build-out without leaning only on debt.
Why Meta Is Raising $95B
Meta's $95 billion plan has two parts: roughly $35 billion in immediate offerings, and a separate $60 billion at-the-market share program that begins by early 2027. The first phase was originally about $22.5 billion in Class A common stock and $12.5 billion in Series A mandatory convertible preferred stock.
Demand allowed Meta to upsize. It raised the common stock offering to $25 billion from $22.5 billion, and with the depositary share offering the raise can reach up to $37.5 billion. Common stock is priced at $625 per share and depositary shares at $50 each (Reuters via Investing.com, October 27, 2026).
Analysts read the raise as a financing fix. Wells Fargo said it likely fulfills Meta's expected $60 billion to $80 billion fiscal 2027 financing gap and reduces reliance on "an increasingly volatile credit market," while Mizuho said it helps remove part of the equity overhang and pulls forward a fix investors expected in January 2027 (Reuters via Investing.com, October 27, 2026).
What the Nvidia Stake Signals
The $3 billion is small against Meta's capital plans, but the pairing with hardware is the point. Meta will deploy more Nvidia Vera Rubin systems, Arm-based Vera CPUs and networking including Spectrum-X Ethernet and BlueField-4, targeting better performance per watt and lower total cost of ownership for Muse training and inference. This expands a multiyear agreement covering millions of Blackwell and Rubin GPUs.
Meta is not locking itself to one supplier. It also works with AMD, builds custom chips, uses external cloud providers, and is considering selling excess compute capacity. That Nvidia, AMD and Broadcom each signaled interest in the offering shows how suppliers are competing for position inside the same customer.
Muse and What Meta Is Funding
Muse is a model family distinct from Llama. Muse Spark launched in April as a multimodal reasoning model, and Muse launched in September as a personal agent that can book travel, fill forms and make purchases. Mark Zuckerberg said, "Muse is how we bring personal superintelligence to everyone" (Techzine, October 27, 2026).
Meta says the proceeds will support its core business and emerging AI opportunities, including its app family, Muse and the infrastructure build-out . The partnership also covers agent security, with Meta considering Nvidia's OpenShell Reference Architecture to deploy agents safely and monitor their behavior.
Risks and Open Questions
Several terms were not disclosed in the coverage reviewed: the mandatory convertible terms, the share count behind Nvidia's $3 billion, and the stake it represents. The size of the second phase, $60 billion through an at-the-market program in 2027, also means Meta's share supply will keep growing after this offering.
The financing depends on spending that has to pay back. With capital expenditures at $135 billion to $145 billion this year, investors will watch whether Muse adoption and Meta's core business generate returns on that scale.
A Market Signal
The Nvidia-Meta deal shows AI infrastructure financing moving into public equity, with suppliers participating as investors alongside their commercial contracts. For fund managers, it sets a reference point for how large AI buyers are raising capital as debt markets tighten.
Dakota Private Markets
Dakota Private Markets is a private markets intelligence and fund performance platform used by investment consultants, allocators, and GP teams to evaluate manager track records, benchmark funds, and identify funds currently raising capital. It provides performance data (Net IRR, TVPI, DPI, and RVPI) for more than 18,000 named funds. Over 1,300 global investment firms use Dakota daily.
Request Access to Dakota Private Markets!
.jpg)
